Comprehensive comparison of two models evaluating eco-environmental quality in Fangshan

It is crucial to employ scientifically sound models for assessing the quality of the ecological environment and revealing the strengths and weaknesses of ecosystems. This process is vital for identifying regional ecological and environmental issues and devising relevant protective measures. Among the widely acknowledged models for evaluating ecological quality, the ecological index (EI) and remote sensing ecological index (RSEI) stand out; however, there is a notable gap in the literature discussing their differences, characteristics, and reasons for selecting either model. In this study, we focused on Fangshan District, Beijing, China, to examine the differences between the two models from 2017 to 2021. We summarized the variations in evaluation indices, importance, quantitative methods, and data acquisition times, proposing application scenarios for both models. The results indicate that the ecological environment quality in Fangshan District, Beijing, remained favorable from 2017 to 2021. There was a discernible trend of initially declining quality followed by subsequent improvement. The variation in the calculation results is evident in the overall correlation between the RSEI and EI. Particularly noteworthy is the significantly smaller correlation between EI and the RSEI in 2021 than in the other two years. This discrepancy is attributed to shifts in the contribution of the evaluation indices within the RSEI model. The use of diverse quantitative methods for evaluating indicators has resulted in several variations. Notably, the evaluation outcomes of the EI model exhibit a stronger correlation with land cover types. This correlation contributes to a more pronounced fluctuation in RSEI levels from 2017 to 2021, with the EI model's evaluation results in 2019 notably surpassing those of the RSEI model. Ultimately, the most prominent disparities lie in the calculation results for water areas and construction land. The substantial difference in water areas is attributed to the distinct importance assigned to evaluation indicators between the two models. Moreover, the notable difference in construction land arises from the use of different quantification methods for evaluation indicators. In general, the EI model has suggested to be more comprehensive and effectively captures the annual comprehensive status of the ecological environment and the multiyear change characteristics of the administrative region. On the other hand, RSEI models exhibit greater flexibility and ease of implementation, independent of spatial and temporal scales. These findings contribute to a clearer understanding of the models' advantages and limitations, offering guidance for decision makers and valuable insights for the improvement and development of ecological environmental quality evaluation models.

High-performance prediction of epilepsy surgical outcomes based on the genetic neural networks and hybrid iEEG marker

Accurately identification of the seizure onset zone (SOZ) is pivotal for successful surgery in patients with medically refractory epilepsy. The purpose of this study is to improve the performance of model predicting the epilepsy surgery outcomes using genetic neural network (GNN) model based on a hybrid intracranial electroencephalography (iEEG) marker. We extracted 21 SOZ related markers based on iEEG data from 79 epilepsy patients. The least absolute shrinkage and selection operator (LASSO) regression was employed to integrated seven markers, selected after testing in pairs with all 21 biomarkers and 7 machine learning models, into a hybrid marker. Based on the hybrid marker, we devised a GNN model and compared its predictive performance for surgical outcomes with six other mainstream machine-learning models. Compared to the mainstream models, underpinning the GNN with the hybrid iEEG marker resulted in a better prediction of surgical outcomes, showing a significant increase of the prediction accuracy from approximately 87% to 94.3% (P = 0.0412). This study suggests that the hybrid iEEG marker can improve the performance of model predicting the epilepsy surgical outcomes, and validates the effectiveness of the GNN in characterizing and analyzing complex relationships between clinical data variables.

Inferring the age and sex of ancient potters from fingerprint ridge densities: A data-driven, Bayesian mixture modelling approach

The density of epidermal ridges in a fingerprint varies predictably by age and sex. Archaeologists are therefore interested in using recovered fingerprints to learn about the ancient people who produced them. Recent studies focus on estimating the age and sex of individuals by measuring their fingerprints with one of two similar metrics: mean ridge breadth (MRB) or ridge density (RD). Yet these attempts face several critical problems: expected values for adult females and adolescent males are inherently indistinguishable, and inter-assemblage variation caused by biological and technological differences cannot be easily estimated. Each of these factors greatly decreases the accuracy of predictions based on individual prints, and together they condemn this strategy to relative uselessness. However, information in fingerprints from across an assemblage can be pooled to generate a more accurate depiction of potter demographics. We present a new approach to epidermal ridge density analysis using Bayesian mixture models with the following key benefits:•Age and sex are estimated more accurately than existing methods by incorporating a data-driven understanding of how demographics and ridge density covary.•Uncertainty in demographic estimates is automatically quantified and included in output.•The Bayesian framework can be easily adapted to fit the unique needs of different researchers.

A quantitative assessment of the consistency of projections from five mathematical models of the HIV epidemic in South Africa: a model comparison study

BACKGROUND

Mathematical models are increasingly used to inform HIV policy and planning. Comparing estimates obtained using different mathematical models can test the robustness of estimates and highlight research gaps. As part of a larger project aiming to determine the optimal allocation of funding for HIV services, in this study we compare projections from five mathematical models of the HIV epidemic in South Africa: EMOD-HIV, Goals, HIV-Synthesis, Optima, and Thembisa.

METHODS

The five modelling groups produced estimates of the total population, HIV incidence, HIV prevalence, proportion of people living with HIV who are diagnosed, ART coverage, proportion of those on ART who are virally suppressed, AIDS-related deaths, total deaths, and the proportion of adult males who are circumcised. Estimates were made under a "status quo" scenario for the period 1990 to 2040. For each output variable we assessed the consistency of model estimates by calculating the coefficient of variation and examining the trend over time.

RESULTS

For most outputs there was significant inter-model variability between 1990 and 2005, when limited data was available for calibration, good consistency from 2005 to 2025, and increasing variability towards the end of the projection period. Estimates of HIV incidence, deaths in people living with HIV, and total deaths displayed the largest long-term variability, with standard deviations between 35 and 65% of the cross-model means. Despite this variability, all models predicted a gradual decline in HIV incidence in the long-term. Projections related to the UNAIDS 95-95-95 targets were more consistent, with the coefficients of variation below 0.1 for all groups except children.

CONCLUSIONS

While models produced consistent estimates for several outputs, there are areas of variability that should be investigated. This is important if projections are to be used in subsequent cost-effectiveness studies.

Effects of blocked versus interleaved training on relative value learning

In reinforcement learning tasks, people learn the values of options relative to other options in the local context. Prior research suggests that relative value learning is enhanced when choice contexts are temporally clustered in a blocked sequence compared to a randomly interleaved sequence. The present study was aimed at further investigating the effects of blocked versus interleaved training using a choice task that distinguishes among different contextual encoding models. Our results showed that the presentation format in which contexts are experienced can lead to qualitatively distinct forms of relative value learning. This conclusion was supported by a combination of model-free and model-based analyses. In the blocked condition, choice behavior was most consistent with a reference point model in which outcomes are encoded relative to a dynamic estimate of the contextual average reward. In contrast, the interleaved condition was best described by a range-frequency encoding model. We propose that blocked training makes it easier to track contextual outcome statistics, such as the average reward, which may then be used to relativize the values of experienced outcomes. When contexts are interleaved, range-frequency encoding may serve as a more efficient means of storing option values in memory for later retrieval.

Individual differences in time-varying and stationary brain connectivity during movie watching from childhood to early adulthood: Age, sex, and behavioral associations

Spatially remote brain regions exhibit dynamic functional interactions across various task conditions. While time-varying functional connectivity during movie watching shows sensitivity to movie content, stationary functional connectivity remains relatively stable across videos. These findings suggest that dynamic and stationary functional interactions may represent different aspects of brain function. However, the relationship between individual differences in time-varying and stationary connectivity and behavioral phenotypes remains elusive. To address this gap, we analyzed an open-access functional MRI dataset comprising participants aged 5-22 years, who watched two cartoon movie clips. We calculated regional brain activity, time-varying connectivity, and stationary connectivity, examining associations with age, sex, and behavioral assessments. Model comparison revealed that time-varying connectivity was more sensitive to age and sex effects compared with stationary connectivity. The preferred age models exhibited quadratic log age or quadratic age effects, indicative of inverted-U shaped developmental patterns. In addition, females showed higher consistency in regional brain activity and time-varying connectivity than males. However, in terms of behavioral predictions, only stationary connectivity demonstrated the ability to predict full-scale intelligence quotient. These findings suggest that individual differences in time-varying and stationary connectivity may capture distinct aspects of behavioral phenotypes.

Value-driven attention and associative learning models: a computational simulation analysis

Value-driven attentional capture (VDAC) refers to a phenomenon by which stimulus features associated with greater reward value attract more attention than those associated with smaller reward value. To date, the majority of VDAC research has revealed that the relationship between reward history and attentional allocation follows associative learning rules. Accordingly, a mathematical implementation of associative learning models and multiple comparison between them can elucidate the underlying process and properties of VDAC. In this study, we implemented the Rescorla-Wagner, Mackintosh (Mac), Schumajuk-Pearce-Hall (SPH), and Esber-Haselgrove (EH) models to determine whether different models predict different outcomes when critical parameters in VDAC were adjusted. Simulation results were compared with experimental data from a series of VDAC studies by fitting two key model parameters, associative strength (V) and associability (α), using the Bayesian information criterion as a loss function. The results showed that SPH-V and EH- α outperformed other implementations of phenomena related to VDAC, such as expected value, training session, switching (or inertia), and uncertainty. Although V of models were sufficient to simulate VDAC when the expected value was the main manipulation of the experiment, α of models could predict additional aspects of VDAC, including uncertainty and resistance to extinction. In summary, associative learning models concur with the crucial aspects of behavioral data from VDAC experiments and elucidate underlying dynamics including novel predictions that need to be verified.

Mapping visual working memory models to a theoretical framework

The body of research on visual working memory (VWM)-the system often described as a limited memory store of visual information in service of ongoing tasks-is growing rapidly. The discovery of numerous related phenomena, and the many subtly different definitions of working memory, signify a challenge to maintain a coherent theoretical framework to discuss concepts, compare models and design studies. A lack of robust theory development has been a noteworthy concern in the psychological sciences, thought to be a precursor to the reproducibility crisis (Oberauer & Lewandowsky, Psychonomic Bulletin & Review, 26, 1596-1618, 2019). I review the theoretical landscape of the VWM field by examining two prominent debates-whether VWM is object-based or feature-based, and whether discrete-slots or variable-precision best describe VWM limits. I share my concerns about the dualistic nature of these debates and the lack of clear model specification that prevents fully determined empirical tests. In hopes of promoting theory development, I provide a working theory map by using the broadly encompassing memory for latent representations model (Hedayati et al., Nature Human Behaviour, 6, 5, 2022) as a scaffold for relevant phenomena and current theories. I illustrate how opposing viewpoints can be brought into accordance, situating leading models of VWM to better identify their differences and improve their comparison. The hope is that the theory map will help VWM researchers get on the same page-clarifying hidden intuitions and aligning varying definitions-and become a useful device for meaningful discussions, development of models, and definitive empirical tests of theories.

An interpretable measure of semantic similarity for predicting eye movements in reading

Predictions about upcoming content play an important role during language comprehension and processing. Semantic similarity as a metric has been used to predict how words are processed in context in language comprehension and processing tasks. This study proposes a novel, dynamic approach for computing contextual semantic similarity, evaluates the extent to which the semantic similarity measures computed using this approach can predict fixation durations in reading tasks recorded in a corpus of eye-tracking data, and compares the performance of these measures to that of semantic similarity measures computed using the cosine and Euclidean methods. Our results reveal that the semantic similarity measures generated by our approach are significantly predictive of fixation durations on reading and outperform those generated by the two existing approaches. The findings of this study contribute to a better understanding of how humans process words in context and make predictions in language comprehension and processing. The effective and interpretable approach to computing contextual semantic similarity proposed in this study can also facilitate further explorations of other experimental data on language comprehension and processing.

Evaluation of semi-distributed hydrological models performance in borkena watershed; upper awash basin, Ethiopia

Flood is one of the most significant disasters in human life and economic destruction. To challenge this disaster, the use of models is very important to predict the magnitude and impact of river flow and to find a solution of the problems. This research is aimed to compare the performance of semi-distributed hydrological models in the Borkena watershed. The selected semi-distributed hydrological models were soil and water assessment tool (SWAT), hydrological engineering center-hydrologic modeling system (HEC-HMS), hydrologiska byråns vattenbalansavdelnin (HBV), and parameter efficient distribution (PED). The models were calibrated from 1999 to 2009 and validated from 2010 to 2015 using daily data. Based on validation results; The Nashsutclif (NSC) output of the SWAT, HEC-HMS, HBV, and PED models were 0.68. 0.66, 0.65, and 0.65, coefficient of determination (R²) 0.69, 0.67, 0.71, and 0.70, percentage of bias (PBIAS) -6.5, 0.6, 27.34, and 10.28, and root mean square error (RMSE) 14.24, 17.45, 17.63 and 0.91, respectively. Based on the models' performance results in Borkena watershed, the first effective model was SWAT and the second one was HEC-HMS. The HBV and PED models took third and fourth places respectively. The overall results show that the two infiltration excess models (SWAT and HEC-HMS) were performed in a better way than the two saturation excess models (HBV and PED) on this watershed. Therefore, according to the model output, the Borkena watershed is an infiltration excess area.

Comparison and multi-model inference of excess risks models for radiation-related solid cancer

In assessments of detrimental health risks from exposures to ionising radiation, many forms of risk to dose-response models are available in the literature. The usual practice is to base risk assessment on one specific model and ignore model uncertainty. The analysis illustrated here considers model uncertainty for the outcome all solid cancer incidence, when modelled as a function of colon organ dose, using the most recent publicly available data from the Life Span Study on atomic bomb survivors of Japan. Seven recent publications reporting all solid cancer risk models currently deemed plausible by the scientific community have been included in a model averaging procedure so that the main conclusions do not depend on just one type of model. The models have been estimated with different baselines and presented for males and females at various attained ages and ages at exposure, to obtain specially computed model-averaged Excess Relative Risks (ERR) and Excess Absolute Risks (EAR). Monte Carlo simulated estimation of uncertainty on excess risks was accounted for by applying realisations including correlations in the risk model parameters. Three models were found to weight the model-averaged risks most strongly depending on the baseline and information criteria used for the weighting. Fitting all excess risk models with the same baseline, one model dominates for both information criteria considered in this study. Based on the analysis presented here, it is generally recommended to take model uncertainty into account in future risk analyses.

New Frontiers in Prevention Research Models: Commentary on the Special Issue

Models considered in the current special issue represent valuable additions to the statistical toolbox of prevention researchers for many types of research questions and designs. Their appropriate use, however, depends on critical evaluation relative to previously existing techniques. This evaluation includes (a.) model choice involving "right-sizing" of the model relative to the amount and quality of data at hand, (b.) examination of the external validity of identified associations relative to observed or latent subgroups, (c.) confirmation of the reasonableness of the functional form assumed by the model, and (d.) identification of influential or outlying observations which unduly affect model fit or parameter estimates. Models in this issue allow for testing of new types of hypotheses in prevention research, and can constitute counterarguments to existing statistical practice. These models may, however, in turn be the object of critical examination of counterarguments a reasonable skeptic may offer.

Comparison of model predictions of typhoid conjugate vaccine public health impact and cost-effectiveness

Models are useful to inform policy decisions on typhoid conjugate vaccine (TCV) deployment in endemic settings. However, methodological choices can influence model-predicted outcomes. To provide robust estimates for the potential public health impact of TCVs that account for structural model differences, we compared four dynamic and one static mathematical model of typhoid transmission and vaccine impact. All models were fitted to a common dataset of age-specific typhoid fever cases in Kolkata, India. We evaluated three TCV strategies: no vaccination, routine vaccination at 9 months of age, and routine vaccination at 9 months with a one-time catch-up campaign (ages 9 months to 15 years). The primary outcome was the predicted percent reduction in symptomatic typhoid cases over 10 years after vaccine introduction. For three models with economic analyses (Models A-C), we also compared the incremental cost-effectiveness ratios (ICERs), calculated as the incremental cost (US$) per disability-adjusted life-year (DALY) averted. Routine vaccination was predicted to reduce symptomatic cases by 10-46 % over a 10-year time horizon under an optimistic scenario (95 % initial vaccine efficacy and 19-year mean duration of protection), and by 2-16 % under a pessimistic scenario (82 % initial efficacy and 6-year mean protection). Adding a catch-up campaign predicted a reduction in incidence of 36-90 % and 6-35 % in the optimistic and pessimistic scenarios, respectively. Vaccine impact was predicted to decrease as the relative contribution of chronic carriers to transmission increased. Models A-C all predicted routine vaccination with or without a catch-up campaign to be cost-effective compared to no vaccination, with ICERs varying from $95-789 per DALY averted; two models predicted the ICER of routine vaccination alone to be greater than with the addition of catch-up campaign. Despite differences in model-predicted vaccine impact and cost-effectiveness, routine vaccination plus a catch-up campaign is likely to be impactful and cost-effective in high incidence settings such as Kolkata.

Digital mapping of soil pH and carbonates at the European scale using environmental variables and machine learning

Soil pH and carbonates (CaCO₃) are important indicators of soil chemistry and fertility, and the prediction of their spatial distribution is critical for the agronomic and environmental management. Digital soil mapping (DSM) techniques are widely accepted for the geospatial analysis of the soil properties. They are rapid and cost-efficient approaches that can provide quantitative prediction. However, the digital mapping of soil pH and CaCO₃ are not well studied, especially at a continental scale. In this research, we mapped the soil pH and CaCO₃ at the European scale using multisource environmental variables and machine learning approaches. Moderate Resolution Imaging Spectroradiometer (MODIS) products, terrain attributes, and climatic variables were considered. Meanwhile, nine machine learning algorithms, namely, three linear and six nonlinear models, were used for the spatial prediction of soil pH and CaCO₃. The land use and cover area frame statistical survey (LUCAS) 2015 topsoil dataset provided by the European Soil Data Centre was utilised. The performances of different models were compared and analysed in terms of coefficient of determination (R²), root mean square error (RMSE), and ratio of performance to deviation (RPD). Specifically, nonlinear machine learning models outperformed the linear ones, and extremely randomized trees (ERT) gave the most satisfactory result for soil pH (R² = 0.70, RMSE = 0.75, and RPD = 1.84) and CaCO₃ (R² = 0.53, RMSE = 93.49 g/kg, and RPD = 1.46). The results revealed that MODIS products and climatic variables were important in predicting soil pH and CaCO₃. Moreover, spatial distribution of soil pH and CaCO₃ in Europe were mapped at 250 m resolution, and the areas with high CaCO₃ content always showed high soil pH value.

Integrating Categorization and Decision-Making

Though individual categorization or decision processes have been studied separately in many previous investigations, few studies have investigated how they interact by using a two-stage task of first categorizing and then deciding. To address this issue, we investigated a categorization-decision task in two experiments. In both, participants were shown six faces varying in width, first asked to categorize the faces, and then decide a course of action for each face. Each experiment was designed to include three groups, and for each group, we manipulated the probabilistic contingencies between stimulus, category assignments, and decision consequences. For each group, each participant received three different sequences of category response, category feedback, decision response, and decision feedback. We found that participants were only partially responsive in the appropriate directions to the contingencies assigned to each group. Comparisons of results from different sequences provided evidence for empirical interference effects of categorization on decisions. The empirical interference effect is defined as the difference between the probability of taking a hostile action in decision-alone conditions and the total probability of taking a hostile action in categorization-decision conditions. To test competing accounts for multiple empirical results, including two-stage choice probabilities and empirical interference effects, we compared a quantum cognition model versus a two-stage exemplar categorization model at both aggregate and individual levels. Using a Bayesian information criterion, we found that the quantum model provided an overall better model fit than the exemplar model. Although both models predicted empirical interference effects, the exemplar model was able to generate probabilistic deviation by incorporating category information of the first stage into the feature representation of the subsequent decision stage, while the quantum model produced interference effect by superposition, measurement, and quantum entanglement.

A comparative study of invariant-based hyperelastic models for silicone elastomers under biaxial deformation with the virtual fields method

Silicone elastomers have been widely used for biomedical applications. A variety of hyperelastic models have been proposed to describe this type of materials in the past few decades. The assessment of the quality of the proposed models is mostly based on stress-strain data obtained from uniform deformation, but very little work has been done to investigate model performances with heterogeneous deformation fields and full-field characterization methods. In this study, thirteen hyperelastic models are evaluated using the virtual fields method combined with full-field deformation data obtained from biaxial tests. The quality of these models is assessed by their capabilities to predict the mechanical responses of silicone elastomers, and the influences of the first and second invariants on modeling of elastomers are investigated through comparative studies between models. The results indicate that for elastomers under finite biaxial deformation, Yeoh model performs the best among selected models; the first invariant plays an important role in constitutive modeling; the second invariant does not have obvious influence on improving the fitting performance. This study provides a full-field method to calibrate and compare hyperelastic models of silicone elastomers under biaxial loading conditions.

Exploring the mediating role of government-public relationships during the COVID-19 pandemic: A model comparison approach

This study proposed, tested, and compared three models to examine an antecedent and outcome of government-public relationships. It conducted three surveys of 9675 people in mainland China, Taiwan, and Hong Kong from August 2020 to January 2021. The results of the model comparison supported the proposed reciprocal model: not only were relational satisfaction and relational trust found to mediate the effect of perceived responsiveness on people's word-of-mouth intention to vaccinate, but they also had a reciprocal influence on each other. This study further affirmed that the relative effects between satisfaction and trust. We also found that emotion-dominant model is more powerful than cognition-dominant model, i.e., people's feeling of satisfaction happens before sense of trust, which results from their perceived organizational responsiveness and then contribute to their word-of-mouth behavioral intention. The theoretical and practical implications of this study were also discussed.

Multiple partition Markov model for B.1.1.7, B.1.351, B.1.617.2, and P.1 variants of SARS-CoV 2 virus

With tools originating from Markov processes, we investigate the similarities and differences between genomic sequences in FASTA format coming from four variants of the SARS-CoV 2 virus, B.1.1.7 (UK), B.1.351 (South Africa), B.1.617.2 (India), and P.1 (Brazil). We treat the virus' sequences as samples of finite memory Markov processes acting in A = { a , c , g , t } . We model each sequence, revealing some heterogeneity between sequences belonging to the same variant. We identified the five most representative sequences for each variant using a robust notion of classification, see Fernández et al. (Math Methods Appl Sci 43(13):7537-7549. 10.1002/mma.5705 ). Using a notion derived from a metric between processes, see García et al. (Appl Stoch Models Bus Ind 34(6):868-878. 10.1002/asmb.2346), we identify four groups, each group representing a variant. It is also detected, by this metric, global proximity between the variants B.1.351 and B.1.1.7. With the selected sequences, we assemble a multiple partition model, see Cordeiro et al. (Math Methods Appl Sci 43(13):7677-7691. 10.1002/mma.6079), revealing in which states of the state space the variants differ, concerning the mechanisms for choosing the next element in A. Through this model, we identify that the variants differ in their transition probabilities in eleven states out of a total of 256 states. For these eleven states, we reveal how the transition probabilities change from variant (group of variants) to variant (group of variants). In other words, we indicate precisely the stochastic reasons for the discrepancies.

The quest for simplicity in human learning: Identifying the constraints on attention

For better or worse, humans live a resource-constrained existence; only a fraction of physical sensations ever reach conscious awareness, and we store a shockingly small subset of these experiences in memory for later use. Here, we examined the effects of attention constraints on learning. Among models that frame selective attention as an optimization problem, attention orients toward information that will reduce errors. Using this framing as a basis, we developed a suite of models with a range of constraints on the attention available during each learning event. We fit these models to both choice and eye-fixation data from four benchmark category-learning data sets, and choice data from another dynamic categorization data set. We found consistent evidence for computations we refer to as "simplicity", where attention is deployed to as few dimensions of information as possible during learning, and "competition", where dimensions compete for selective attention via lateral inhibition.

The Model Dimensionality and Its Impacts on the Strategic and Policy Outcomes in IAMs the Findings from the RICE2020 Model

This paper studies the impacts of regional breakdowns or model dimensionality on the model's optimal solutions. Using the United States (USA) and China (CHN) as the experimental subject, we test the various solutions related to USA and CHN, such as the Cournot-Nash equilibrium and the Lindahl equilibrium, in the RICE2020 model under three regional breakdowns. Their solutions' invariance and variances across different model dimensionalities indicate that modeling dimensionality may play a role in the strategic interactions among the regions in GHG mitigation. The simulation results also point out the pitfalls of the model comparisons across IAMs for climate change.

Estimation of COVID-19 mortality in the United States using Spatio-temporal Conway Maxwell Poisson model

Spatio-temporal Poisson models are commonly used for disease mapping. However, after incorporating the spatial and temporal variation, the data do not necessarily have equal mean and variance, suggesting either over- or under-dispersion. In this paper, we propose the Spatio-temporal Conway Maxwell Poisson model. The advantage of Conway Maxwell Poisson distribution is its ability to handle both under- and over-dispersion through controlling one special parameter in the distribution, which makes it more flexible than Poisson distribution. We consider data from the pandemic caused by the SARS-CoV-2 virus in 2019 (COVID-19) that has threatened people all over the world. Understanding the spatio-temporal pattern of the disease is of great importance. We apply a spatio-temporal Conway Maxwell Poisson model to data on the COVID-19 deaths and find that this model achieves better performance than commonly used spatio-temporal Poisson model.

Anxiety and depression related to childhood maltreatment in teenagers: Comparing multiple individual risk model, cumulative risk model and latent profile analysis

BACKGROUND

Few studies have employed person-centered approaches (i.e. latent profile analysis in this study) to investigate the specific patterns of childhood maltreatment in a large sample of Chinese adolescents, and little is known about the predictive validity of latent profile analysis on internalizing problems, compared with multiple individual risk model and cumulative risk model.

OBJECTIVE

The purpose of this study was to investigate whether differential patterns of maltreatment existed by employing latent profile analysis with a sample of 9071 Chinese adolescents, and further examined the predictive validity of latent profile analysis on internalizing problems, relative to the cumulative risk and multiple individual risk model.

PARTICIPANTS

Using a stratified sampling approach, 10,515 participants (Mean age = 14.24; SD = 1.73) were chosen from three different types of middle schools in Chongqing city, China. 9071 valid responses (males = 4775; females = 4296) were obtained for final analysis.

METHODS

Participants reported their childhood maltreatment experience, anxiety and depression symptoms. Latent profile analysis was used to obtain possible patterns of maltreatment with Mplus version 7. 4. Bolck-Croon-Hagenaars (BCH) method was used to test the association between maltreatment patterns and anxiety and depression symptoms. Relative weight analysis and analysis of variance (ANOVA) were used to test the predictive validity of latent profile analysis, multiple individual risk and cumulative risk model.

RESULTS

Using latent profile analysis, two patterns of childhood maltreatment were uncovered ("No Maltreatment" and "Multiple Maltreatment"). Further analysis showed that multiple individual risk model accounted for the largest variance in anxiety (R² = 26.7%) and depression (R² = 33%), followed by the latent profile analysis (R² = 14.7% for anxiety and 18.6% for depression) and the cumulative risk model (R² = 12.9% for anxiety and 15.2% for depression).

CONCLUSIONS

Our findings suggested that the multiple individual risk model is the optimal model for identifying adolescents at the risk of developing anxiety and depression symptoms, and the results suggested emotional abuse and emotional neglect are risk factors for higher levels of anxiety and depression among adolescents.

Improving monitoring of fish health in the oil sands region using regularization techniques and water quality variables

Trout-perch are sampled from the Athabasca River in Alberta, Canada, as a sentinel species for environmental health. The performance of trout-perch populations is known to be influenced by the quality of the water in which they reside. Using climate, environmental, and water quality variables measured in the Athabasca River near trout-perch sampling locations is found to improve model fitting and the predictability of models for the adjusted body weight, adjusted gonad weight, and adjusted liver weight of trout-perch. Given a large number of covariables, three variable selection techniques: stepwise regression, the lasso, and the elastic net (EN) are considered for selecting a subset of relevant variables. The models selected by the lasso and EN are found to outperform the models selected by stepwise regression in general, and little difference is observed between the models selected by the lasso and EN. Uranium, tungsten, tellurium, pH, molybdenum, and antimony are selected for at least one fish response.

Loneliness and functional limitations among older adults with diabetes: Comparing directional models

OBJECTIVE

Middle-aged and older adults with diabetes are at increased risk for loneliness and functional limitations. Cross-sectional and longitudinal associations between loneliness and functional limitations have been demonstrated among the general population, but have not been established among those with diabetes. The purpose of this study was to directly compare the following models describing the direction of the association between loneliness and functional limitations among people with diabetes: (1) loneliness leads to functional limitations, (2) functional limitations lead to loneliness, and (3) a bidirectional association between loneliness and functional limitations.

METHODS

Data came from the Health and Retirement Study. Participants were middle-aged and older individuals with diabetes in the United States (n = 2934). Loneliness and functional limitations were measured at baseline, 4-year follow-up, and 8-year follow-up. Path models for each of the three models, as well as a stability model, were created. Model fit was compared using Akaike's Information Criteria (AIC).

RESULTS

Participants were 54.6% female, 74.98% White, had a mean age of 69.66 years, had an average of 1.48 comorbid chronic conditions, and had diabetes for an average of 10.40 years. The bidirectional model best fit the data as evidenced by the lowest AIC value (AIC = 171,162.81). ∆AIC between the bidirectional model and the next best fitting model was 16.19, indicating strong support for selecting the bidirectional model. Higher levels of loneliness were associated with subsequent higher levels of functional limitations at some time points (βs = 0.07, 0.02) and higher levels of functional limitations were associated with subsequent higher levels of loneliness (βs = 0.13, 0.06) at all time points.

CONCLUSION

Results suggest that the association between loneliness and functional limitations among individuals with diabetes is bidirectional. This study demonstrates the value of directly comparing directional models.

An introduction to thermodynamic integration and application to dynamic causal models

In generative modeling of neuroimaging data, such as dynamic causal modeling (DCM), one typically considers several alternative models, either to determine the most plausible explanation for observed data (Bayesian model selection) or to account for model uncertainty (Bayesian model averaging). Both procedures rest on estimates of the model evidence, a principled trade-off between model accuracy and complexity. In the context of DCM, the log evidence is usually approximated using variational Bayes. Although this approach is highly efficient, it makes distributional assumptions and is vulnerable to local extrema. This paper introduces the use of thermodynamic integration (TI) for Bayesian model selection and averaging in the context of DCM. TI is based on Markov chain Monte Carlo sampling which is asymptotically exact but orders of magnitude slower than variational Bayes. In this paper, we explain the theoretical foundations of TI, covering key concepts such as the free energy and its origins in statistical physics. Our aim is to convey an in-depth understanding of the method starting from its historical origin in statistical physics. In addition, we demonstrate the practical application of TI via a series of examples which serve to guide the user in applying this method. Furthermore, these examples demonstrate that, given an efficient implementation and hardware capable of parallel processing, the challenge of high computational demand can be overcome successfully. The TI implementation presented in this paper is freely available as part of the open source software TAPAS.

Comparing traditional modeling approaches versus predictive analytics methods for predicting multiple sclerosis relapse

OBJECTIVE

This study compared traditional statistical methods to different predictive analytics methods on the endpoint of multiple sclerosis (MS) relapse.

STUDY SETTING

This is a secondary data analysis on four different MS Centers based on the third year of data, July 2019-June 2020.

STUDY DESIGN

The parent study is a two-part, 3-year clinical quality improvement prospective study that started in June 2017 and concluded in June 2020, and utilizes a prospective stepped-wedge randomized design. Binary logistic regression was compared with other machine learning models, specifically ridge, least absolute shrinkage and selection operator (LASSO), and random forest.

DATA COLLECTION

This study used electronic health record data extracted at the individual level and 'rolled up' to the system and population level. Inclusion criteria included participants aged 18 years or older, with MS presenting to any of the four centers, who entered the study in any quarter. Exclusion criteria included cases with missing or incorrectly input data and those who refused to participate in the study.

PRINCIPAL FINDINGS

When comparing relapse indices across models, random forest significantly outperformed logistic regression and other machine learning algorithms (Δperf_A =27.1%, Δperf_M =27.5%). However, for Δperf_F, logistic regression and random forest performed relatively the same. Ridge and LASSO outperformed logistic regression (Δperf_M1 =0.9%, Δperf_M2 =9.4%, Δperf_F2=25.8%, respectively).

CONCLUSION

Multiple sclerosis is a complex and costly chronic ("3C") condition that currently has no cure. In a condition like MS, which has an unpredictable course, the use of predictive analytics could help health systems learn better, faster, and to improve more effectively and predict rather than react to emerging health needs for people with MS. Comparing the predictability of relapse across various models with a predictive analytics framework can potentially change how we manage MS care.

From data compilation to model validation: comparing three ecosystem models of the Tasman and Golden Bays, New Zealand

The Tasman and Golden Bays (TBGB) are a semi-enclosed embayment system in New Zealand that supports numerous commercial and recreational activities. We present three ecosystem models of the TBGB ecosystem with varying levels of complexity, aimed at contributing as tools to aid in understanding this ecosystem and its responses to anthropogenic and natural pressures. We describe the process of data compilation through to model validation and analyse the importance of knowledge gaps with respect to model dynamics and results. We compare responses in all three models to historical fishing, and analyse similarities and differences in the dynamics of the three models. We assessed the most complex of the models against initialisation uncertainty and sensitivity to oceanographic variability and found it most sensitive to the latter. We recommend that scenarios relating to ecosystem dynamics of the TBGB ecosystem include sensitivities, especially oceanographic uncertainty, and compare responses across all three models where it is possible to do so.

Using predicted length of stay to define treatment and model costs in hospitalized adults with serious illness: an evaluation of palliative care

BACKGROUND

Economic research on hospital palliative care faces major challenges. Observational studies using routine data encounter difficulties because treatment timing is not under investigator control and unobserved patient complexity is endemic. An individual's predicted LOS at admission offers potential advantages in this context.

METHODS

We conducted a retrospective cohort study on adults admitted to a large cancer center in the United States between 2009 and 2015. We defined a derivation sample to estimate predicted LOS using baseline factors (N = 16,425) and an analytic sample for our primary analyses (N = 2674) based on diagnosis of a terminal illness and high risk of hospital mortality. We modelled our treatment variable according to the timing of first palliative care interaction as a function of predicted LOS, and we employed predicted LOS as an additional covariate in regression as a proxy for complexity alongside diagnosis and comorbidity index. We evaluated models based on predictive accuracy in and out of sample, on Akaike and Bayesian Information Criteria, and precision of treatment effect estimate.

RESULTS

Our approach using an additional covariate yielded major improvement in model accuracy: R² increased from 0.14 to 0.23, and model performance also improved on predictive accuracy and information criteria. Treatment effect estimates and conclusions were unaffected. Our approach with respect to treatment variable yielded no substantial improvements in model performance, but post hoc analyses show an association between treatment effect estimate and estimated LOS at baseline.

CONCLUSION

Allocation of scarce palliative care capacity and value-based reimbursement models should take into consideration when and for whom the intervention has the largest impact on treatment choices. An individual's predicted LOS at baseline is useful in this context for accurately predicting costs, and potentially has further benefits in modelling treatment effects.

Estimating forest floor carbon stocks in woodland formations in Spain

The forest floor C stock needs to be accurately estimated in order to quantify its contribution to nutrient cycling and other ecological processes as well as for reporting purposes under international agreements. Hence, a modelling approach was used which involved testing three different types of models (GLM, GAM and random forest) to determine which one provided the best estimates of forest floor C stocks. The dataset employed contained over 1650 observations from different available sources embracing different climatic, topographic and biotic variables to be tested in the model. The approach that provided the best estimation of forest floor C stock was the random forest method, with forest type, latitude, altitude, canopy cover, mean summer temperature, annual accumulated temperature, summer precipitation, water deficit and the normalized difference vegetation index (NDVI) as covariates. To obtain a robust forecast, several iterations of the model were performed to estimate forest floor C stocks from the mean of the predictions. The model estimated a forest floor C stock of 0.148 ± 0.081 Pg, equivalent to a biomass of 0.381 ± 0.214 Pg, for a wooded area of almost 184,000 km² in peninsular Spain and the Balearic Islands. The predictions were also presented in the form of a map showing the spatial distribution of the forest floor C stock. The results revealed a mean forest floor C stock of 8 Mg C ha^-1 for Spanish forests and identified differences between coniferous (10.1 Mg C ha^-1) and hardwood forests (6.3 Mg C ha^-1).

Internal-external cross-validation helped to evaluate the generalizability of prediction models in large clustered datasets

OBJECTIVE

To illustrate how to evaluate the need of complex strategies for developing generalizable prediction models in large clustered datasets.

STUDY DESIGN AND SETTING

We developed eight Cox regression models to estimate the risk of heart failure using a large population-level dataset. These models differed in the number of predictors, the functional form of the predictor effects (non-linear effects and interaction) and the estimation method (maximum likelihood and penalization). Internal-external cross-validation was used to evaluate the models' generalizability across the included general practices.

RESULTS

Among 871,687 individuals from 225 general practices, 43,987 (5.5%) developed heart failure during a median follow-up time of 5.8 years. For discrimination, the simplest prediction model yielded a good concordance statistic, which was not much improved by adopting complex strategies. Between-practice heterogeneity in discrimination was similar in all models. For calibration, the simplest model performed satisfactorily. Although accounting for non-linear effects and interaction slightly improved the calibration slope, it also led to more heterogeneity in the observed/expected ratio. Similar results were found in a second case study involving patients with stroke.

CONCLUSION

In large clustered datasets, prediction model studies may adopt internal-external cross-validation to evaluate the generalizability of competing models, and to identify promising modelling strategies.

Elixhauser outperformed Charlson comorbidity index in prognostic value after ACS: insights from a national registry

OBJECTIVE

To compare the performance of risk adjustment models using the Elixhauser and Charlson comorbidity scores in predicting in-hospital outcomes of ACS patients from a nationwide administrative database.

STUDY DESIGN AND SETTING

All hospitalizations for ACS in the United States between 2004 and 2014 (n = 7,201,900) were retrospectively analyzed. We used ECS and CCI score based on ICD-9 codes to define comorbidity variables. Logistic regression models were fitted to three in-hospital outcomes, including mortality, Major Acute Cardiovascular & Cerebrovascular Events (MACCE) and bleeding. The prognostic values of ECS and CCI after adjusting for known confounders, were compared using the C-statistic, Akaike information criterion (AIC), and Bayesian information criterion (BIC).

RESULTS

The statistical performance of models predicting all in-hospital outcomes demonstrated that the ECS had superior prognostic value compared to the CCI, with higher C-statistics and lower AIC and BIC values associated with the former.

CONCLUSION

This is the first study that compared the prognostic value of the ECS and CCI scores in predicting multiple ACS outcomes, based on their scoring systems. Better discrimination and goodness of fit was achieved with the Elixhauser method across all in-hospital outcomes studied.

Simple Threshold Rules Solve Explore/Exploit Trade-offs in a Resource Accumulation Search Task

How, and how well, do people switch between exploration and exploitation to search for and accumulate resources? We study the decision processes underlying such exploration/exploitation trade-offs using a novel card selection task that captures the common situation of searching among multiple resources (e.g., jobs) that can be exploited without depleting. With experience, participants learn to switch appropriately between exploration and exploitation and approach optimal performance. We model participants' behavior on this task with random, threshold, and sampling strategies, and find that a linear decreasing threshold rule best fits participants' results. Further evidence that participants use decreasing threshold-based strategies comes from reaction time differences between exploration and exploitation; however, participants themselves report non-decreasing thresholds. Decreasing threshold strategies that "front-load" exploration and switch quickly to exploitation are particularly effective in resource accumulation tasks, in contrast to optimal stopping problems like the Secretary Problem requiring longer exploration.

A note on identification constraints and information criteria in Bayesian latent variable models

It is well known that, in traditional SEM applications, a scale must be set for each latent variable: typically, either the latent variance or a factor loading is fixed to one. While this has no impact on the fit metrics in ML estimation, it can potentially lead to varying Bayesian model comparison metrics due to the use of different prior distributions under each parameterization. This is a problem, because a researcher could artificially improve one's preferred model simply by changing the identification constraint. Using a single-factor CFA as motivation for study, we first show that Bayesian model comparison metrics can systematically change depending on constraints used. We then study principled methods for setting the scale of the latent variable that stabilize the model comparison metrics. These methods involve (i) the placement of priors on ratios of factor loadings, as opposed to individual loadings; and (ii) use of effect coding. We illustrate the methods via simulation and application.

The scaled target learning model: Revisiting learning in the balloon analogue risk task

The Balloon Analogue Risk Task (BART) is a sequential decision making paradigm that assesses risk-taking behavior. Several computational models have been proposed for the BART that characterize risk-taking propensity. An aspect of task performance that has proven challenging to model is the learning that develops from experiencing wins and losses across trials, which has the potential to provide further insight into risky decision making. We developed the Scaled Target Learning (STL) model for this purpose. STL describes learning as adjustments to an individual's strategy in reaction to outcomes in the task, with the size of adjustments reflecting an individual's sensitivity to wins and losses. STL is shown to be sensitive to the learning elicited by experimental manipulations. In addition, the model matches or bests the performance of three competing models in traditional model comparison tests (e.g., parameter recovery performance, predictive accuracy, sensitivity to risk-taking propensity). Findings are discussed in the context of the learning process involved in the task. By characterizing the extent to which people are willing to adapt their strategies based on past experience, STL is a step toward a complete depiction of the psychological processes underlying sequential risk-taking behavior.

To pool or not to pool: That is the question in microbial kinetics

Variation observed in heat inactivation of Salmonella strains (data from Combase) was characterized using multilevel modeling with two case studies. One study concerned repetitions at one temperature, the other concerned isothermal experiments at various temperatures. Multilevel models characterize variation at various levels and handle dependencies in the data. The Weibull model was applied using Bayesian regression. The research question was how parameters varied with experimental conditions and how data can best be analyzed: no pooling (each experiment analyzed separately), complete pooling (all data analyzed together) or partial pooling (connecting the experiments while allowing for variation between experiments). In the first case study, level 1 consisted of the measurements, level 2 of the group of repetitions. While variation in the initial number parameter was low (set by the researchers), the Weibull shape factor varied for each repetition from 0.58-1.44, and the rate parameter from 0.006-0.074 h. With partial pooling variation was much less, with complete pooling variation was strongly underestimated. In the second case study, level 1 consisted of the measurements, level 2 of the group of repetitions per temperature experiment, level 3 of the cluster of various temperature experiments. The research question was how temperature affected the Weibull parameters. Variation in initial numbers was low (set by the researchers), the rate parameter was obviously affected by temperature, the estimate of the shape parameter depended on how the data were analyzed. With partial pooling, and one-step global modeling with a Bigelow-type model for the rate parameter, shape parameter variation was minimal. Model comparison based on prediction capacity of the various models was explored. The probability distribution of calculated decimal reduction times was much narrower using multilevel global modeling compared to the usual single level two-step approach. Multilevel modeling of microbial heat inactivation appears to be a suitable and powerful method to characterize and quantify variation at various levels. It handles possible dependencies in the data, and yields unbiased parameter estimates. The answer on the question "to pool or not to pool" depends on the goal of modeling, but if the goal is prediction, then partial pooling using multilevel modeling is the answer, provided that the experimental data allow that.

Comparison of different groundwater vulnerability evaluation models of typical karst areas in north China: a case of Hebi City

Groundwater pollution is a serious problem in north China. However, the study on the vulnerability of karst groundwater is mainly in south China, and there are few studies in north China. To study the applicability of different models of karst areas in north China, this paper chose a special study area-Hebi City, where the exposed karst area is widely developed in the hilly area, but the covered karst area is in the eastern part of the study area. The DRASTIC model, the AHP-DRASTIC model, and the improved COPK model were adopted to evaluate the vulnerability of shallow karst groundwater in Hebi City. Cl^-, SO₄^2-, NO₃^-, and TDS were selected to verify the rationality of the evaluation results. It shows that the improved COPK model is more suitable for the shallow karst groundwater vulnerability evaluation in the karst areas in northern China represented by the study area than the other two. The study area was divided into 4 classes by the improved COPK model: highest (14.07%), high (53.05%), low (21.37%), and lowest (11.51%). Then, the analytic hierarchy process and comprehensive index model were used to evaluate the groundwater pollution load intensity, and the study area was divided into 3 classes: high (23.33%), moderate (64.66%), and low (12.01%). According to the analysis of the relationship between groundwater pollution load intensity and groundwater quality, it can be found that human activities have an obvious influence on groundwater quality in the study area. Finally, combined with human activities, the study area was divided into 3 remediation areas, 1 control area, and 1 protected area. This paper can provide a scientific basis for rational exploitation and utilization of groundwater resources. It can also provide a reference for future generations to evaluate the groundwater vulnerability in the northern China karst areas.

Comparison and evaluation of the simulated annual aerosol characteristics over China with two global aerosol models

In this study, simulations of the annual mean aerosol budget, aerosol optical properties, and surface mass concentration in 2006 in China are performed with two aerosol interactive global atmosphere models, namely, the Nonhydrostatic ICosahedral Atmospheric Model (NICAM) coupled with the Spectral Radiation Transport Model for Aerosol Species (SPRINTARS) and the Beijing Climate Center Atmospheric General Circulation Model (BCC_AGCM) coupled with the Canadian Aerosol Module (CAM) online. The observed and simulated aerosol optical depths (AODs) exhibit similar horizontal distributions with large values over eastern and central China, and sulfate aerosols contribute the main differences between the AODs simulated by NICAM and BCC_AGCM. The simulated sulfate and dust surface concentrations are more consistent with observations compared with the simulated carbonaceous surface concentrations, and both models can reproduce the decreasing tendency of the sulfate surface concentration from urban sites to rural sites. However, the dust emission and deposition levels in China simulated by BCC_AGCM are three times as high as those simulated by NICAM, and the major sink processes of the anthropogenic sulfate, black carbon (BC), and organic carbon (OC) aerosols over China are very different between the two models. The emission and deposition results, which are closely related to the model-assumed aerosol particle size distribution, indicate that the current aerosol size distribution used in the two models should be further improved. The differences in dust emission parameterizations also lead significant discrepancies in aerosol cycles and the dust emission scheme is an important factor determining the magnitudes of global and regional dust emission fluxes.

Effect estimates of COVID-19 non-pharmaceutical interventions are non-robust and highly model-dependent

Objective

To compare the inference regarding the effectiveness of the various non-pharmaceutical interventions (NPIs) for COVID-19 obtained from different SIR models.

Study design and setting

We explored two models developed by Imperial College that considered only NPIs without accounting for mobility (model 1) or only mobility (model 2), and a model accounting for the combination of mobility and NPIs (model 3). Imperial College applied models 1 and 2 to 11 European countries and to the USA, respectively. We applied these models to 14 European countries (original 11 plus another 3), over two different time horizons.

Results

While model 1 found that lockdown was the most effective measure in the original 11 countries, model 2 showed that lockdown had little or no benefit as it was typically introduced at a point when the time-varying reproduction number was already very low. Model 3 found that the simple banning of public events was beneficial, while lockdown had no consistent impact. Based on Bayesian metrics, model 2 was better supported by the data than either model 1 or model 3 for both time horizons.

Conclusion

Inferences on effects of NPIs are non-robust and highly sensitive to model specification. In the SIR modeling framework, the impacts of lockdown are uncertain and highly model-dependent.

Improving national level spatial mapping of malaria through alternative spatial and spatio-temporal models

The most common approach to create spatial prediction of malaria in the literature is to approximate a Gaussian process model using stochastic partial differential equation (SPDE). We compared SPDE to computationally faster alternatives, generalized additive model (GAM) and state-of-the-art machine learning method gradient boosted trees (GBM), with respect to their predictive skill for country-level malaria prevalence mapping. We also evaluated the intuition that incorporation of past data and the use of spatio-temporal models may improve predictive accuracy of present spatial distribution of malaria. Model performances varied among the countries and setting with SPDE and GAM performed well generally. The inclusion of past data is beneficial for GAM and GBM, but not for SPDE. We further investigated the weaknesses of SPDE at spatio-temporal setting and GAM at the edges of the countries. Taken together, we believe that spatial/spatio-temporal SPDE models should be evaluated alongside with the alternatives or at least GAM.

Contrasting cumulative risk and multiple individual risk models of the relationship between Adverse Childhood Experiences (ACEs) and adult health outcomes

BACKGROUND

A very large body of research documents relationships between self-reported Adverse Childhood Experiences (srACEs) and adult health outcomes. Despite multiple assessment tools that use the same or similar questions, there is a great deal of inconsistency in the operationalization of self-reported childhood adversity for use as a predictor variable. Alternative conceptual models are rarely used and very limited evidence directly contrasts conceptual models to each other. Also, while a cumulative numeric 'ACE Score' is normative, there are differences in the way it is calculated and used in statistical models. We investigated differences in model fit and performance between the cumulative ACE Score and a 'multiple individual risk' (MIR) model that enters individual ACE events together into prediction models. We also investigated differences that arise from the use of different strategies for coding and calculating the ACE Score.

METHODS

We merged the 2011-2012 BRFSS data (N = 56,640) and analyzed 3 outcomes. We compared descriptive model fit metrics and used Vuong's test for model selection to arrive at best fit models using the cumulative ACE Score (as both a continuous or categorical variable) and the MIR model, and then statistically compared the best fit models to each other.

RESULTS

The multiple individual risk model was a better fit than the categorical ACE Score for the 'lifetime history of depression' outcome. For the outcomes of obesity and cardiac disease, the cumulative risk and multiple individual risks models were of comparable fit, but yield different and complementary inferences.

CONCLUSIONS

Additional information-rich inferences about ACE-health relationships can be obtained from including a multiple individual risk modeling strategy. Results suggest that investigators working with large srACEs data sources could empirically derive the number of items, as well as the exposure coding strategy, that are a best fit for the outcome under study. A multiple individual risk model could also be considered in addition to the cumulative risk model, potentially in place of estimation of unadjusted ACE-outcome relationships.

Modeling socio-demographic and clinical factors influencing psychiatric inpatient service use: a comparison of models for zero-Inflated and overdispersed count data

BACKGROUND

Psychiatric disorders may occur as a single episode or be persistent and relapsing, sometimes leading to suicidal behaviours. The exact causes of psychiatric disorders are hard to determine but easy access to health care services can help to reduce their severity. The aim of this study was to investigate the factors associated with repeated hospitalizations among the patients with psychiatric illness, which may help the policy makers to target the high-risk groups in a more focused manner.

METHODS

A large linked administrative database consisting of 200,537 patients with psychiatric diagnosis in the years of 2008-2012 was used in this analysis. Various counts regression models including zero-inflated and hurdle models were considered for analyzing the hospitalization rate among patients with psychiatric disorders within three months follow-up since their index visit dates. The covariates for this study consisted of socio-demographic and clinical characteristics of the patients.

RESULTS

The results show that the odds of hospitalization are significantly higher among registered Indians, male patients and younger patients. Hospitalization rate depends on the patients' disease types. Having previously visited a general physician served a protective role for psychiatric hospitalization during the study period. Patients who had seen an outpatient psychiatrist were more likely to have a higher number of psychiatric hospitalizations. This may indicate that psychiatrists tend to see patients with more severe illnesses, who require hospital-based care for managing their illness.

CONCLUSIONS

Providing easier access to registered Indian people and youth may reduce the need for hospital-based care. Patients with mental health conditions may benefit from greater and more timely access to primary care.

RBE variation in prostate carcinoma cells in active scanning proton beams: In-vitro measurements in comparison with phenomenological models

PURPOSE

In-vitro radiobiological studies are essential for modelling the relative biological effectiveness (RBE) in proton therapy. The purpose of this study was to experimentally determine the RBE values in proton beams along the beam path for human prostate carcinoma cells (Du-145). RBE-dose and RBE-LET_d (dose-averaged linear energy transfer) dependencies were investigated and three phenomenological RBE models, i.e. McNamara, Rørvik and Wilkens were benchmarked for this cell line.

METHODS

Cells were placed at multiple positions along the beam path, employing an in-house developed solid phantom. The experimental setup reflected the clinical prostate treatment scenario in terms of field size, depth, and required proton energies (127.2-180.1 MeV) and the physical doses from 0.5 to 6 Gy were delivered. The reference irradiation was performed with 200 kV X-ray beams. Respective (α/β) values were determined using the linear quadratic model and LET_d was derived from the treatment planning system at the exact location of cells.

RESULTS AND CONCLUSION

Independent of the cell survival level, all experimental RBE values were consistently higher in the target than the generic clinical RBE value of 1.1; with the lowest RBE value of 1.28 obtained at the beginning of the SOBP. A systematic RBE decrease with increasing dose was observed for the investigated dose range. The RBE values from all three applied models were considerably smaller than the experimental values. A clear increase of experimental RBE values with LET_d parameter suggests that proton LET must be taken into consideration for this low (α/β) tissue.

Longitudinal change in physical functioning and dropout due to death among the oldest old: a comparison of three methods of analysis

Longitudinal studies examining changes in physical functioning with advancing age among very old people are plagued by high death rates, which can lead to biased estimates. This study was conducted to analyse changes in physical functioning among the oldest old with three distinct methods which differ in how they handle dropout due to death. The sample consisted of 3992 persons aged 90 or over in the Vitality 90+ Study who were followed up on average for 2.5 years (range 0–13 years). A generalized estimating equation (GEE) with independent ‘working’ correlation, a linear mixed-effects (LME) model and a joint model consisting of longitudinal and survival submodels were used to estimate the effect of age on physical functioning over 13 years of follow-up. We observed significant age-related decline in physical functioning, which furthermore accelerated significantly with age. The average rate of decline differed markedly between the models: the GEE-based estimate for linear decline among survivors was about one-third of the average individual decline in the joint model and half the decline indicated by the LME model. In conclusion, the three methods yield substantially different views on decline in physical functioning: the GEE model may be useful for considering the effect of intervention measures on the outcome among living people, whereas the LME model is biased regarding studying outcomes associated with death. The joint model may be valuable for predicting the future characteristics of the oldest old and planning elderly care as life expectancy continues gradually to rise.

An Information Theoretic Approach to Model Selection: A Tutorial with Monte Carlo Confirmation

A reliance on null hypothesis significance testing (NHST) and misinterpretations of its results are thought to contribute to the replication crisis while impeding the development of a cumulative science. One solution is a data-analytic approach called Information-Theoretic (I-T) Model Selection, which builds upon Maximum Likelihood estimates. In the I-T approach, the scientist examines a set of candidate models and determines for each one the probability that it is the closer to the truth than all others in the set. Although the theoretical development is subtle, the implementation of I-T analysis is straightforward. Models are sorted according to the probability that they are the best in light of the data collected. It encourages the examination of multiple models, something investigators desire and that NHST discourages. This article is structured to address two objectives. The first is to illustrate the application of I-T data analysis to data from a virtual experiment. A noisy delay-discounting data set is generated and seven quantitative models are examined. In the illustration, it is demonstrated that it is not necessary to know the "truth" is to identify the one that is closest to it and that the most likely models conform to the model that generated the data. Second, we examine claims made by advocates of the I-T approach using Monte Carlo simulations in which 10,000 different data sets are generated and analyzed. The simulations showed that 1) the probabilities associated with each model returned by the single virtual experiment approximated those that resulted from the simulations, 2) models that were deemed close to the truth produced the most precise parameter estimates, and 3) adding a single replicate sharpens the ability to identify the most probable model.

Anticipating the impact of pitfalls in kinetic biodegradation parameter estimation from substrate depletion curves of organic pollutants

Accurate and reliable estimation of kinetic parameters of pollutant biodegradation processes is essential for environmental and health risk assessment. Common biodegradation models proposed in the literature, such as the nonlinear Monod equation and its simplified versions (e.g. Michaelis-Menten-like and first-order equations), are problematic in terms of accuracy of kinetic parameters due to the parameter correlation. However, a comparison between these models in terms of accuracy and reliability, related to data imprecision, has not been performed in the literature. This task is necessary, mainly because the model selection cannot be straightforward, as shown in this work. To facilitate the comparison, novel statistics summarising the accuracy and reliability of estimations are introduced. The main objective is to establish relationships between these statistics (trough new diagnostic indicators) to limit the probability of pitfalls or to avoid the negative impact of an improper model selection. Such anticipation is an imperative need in the biodegradation modelling framework and, to the best of our knowledge, it has never been performed. In order to account for accuracy, simulated data in realistic conditions are used to highlight the magnitude of pitfalls related to the model selection for estimation of the main kinetic parameters (K_s, μ_m and/or V_m). Four scenarios related to model selection are compared for the first time and, diagnostic indicators able to anticipate relevant aspects related to accuracy and reliability are introduced. Moreover, first evidences of the impact of measurement errors in other intrinsic Monod parameters (the initial biomass concentration and the microbial yield coefficient, Y), as well as the impact of the simultaneous μ_m, K_s and Y estimation, on the accuracy and reliability of the estimations are reported. Despite the pitfalls shown, specific applicability of even unreliable models is highlighted, and suggestions for environmental and health risk modellers are provided accordingly.

The c-index is not proper for the evaluation of $t$-year predicted risks

We show that the widely used concordance index for time to event outcome is not proper when interest is in predicting a $t$-year risk of an event, for example 10-year mortality. In the situation with a fixed prediction horizon, the concordance index can be higher for a misspecified model than for a correctly specified model. Impropriety happens because the concordance index assesses the order of the event times and not the order of the event status at the prediction horizon. The time-dependent area under the receiver operating characteristic curve does not have this problem and is proper in this context.

How to model the neurocognitive dynamics of decision making: A methodological primer with ACT-R

Higher cognitive functions are the product of a dynamic interplay of perceptual, mnemonic, and other cognitive processes. Modeling the interplay of these processes and generating predictions about both behavioral and neural data can be achieved with cognitive architectures. However, such architectures are still used relatively rarely, likely because working with them comes with high entry-level barriers. To lower these barriers, we provide a methodological primer for modeling higher cognitive functions and their constituent cognitive subprocesses with arguably the most developed cognitive architecture today-ACT-R. We showcase a principled method of generating individual response time predictions, and demonstrate how neural data can be used to refine ACT-R models. To illustrate our approach, we develop a fully specified neurocognitive model of a prominent strategy for memory-based decisions-the take-the-best heuristic-modeling decision making as a dynamic interplay of perceptual, motor, and memory processes. This implementation allows us to predict the dynamics of behavior and the temporal and spatial patterns of brain activity. Moreover, we show that comparing the predictions for brain activity to empirical BOLD data allows us to differentiate competing ACT-R implementations of take the best.

Two-step actions in infancy-the TWAIN model

In this paper, we propose a novel model—the TWAIN model—to describe the durations of two-step actions in a reach-to-place task in human infants. Previous research demonstrates that infants and adults plan their actions across multiple steps. They adjust, for instance, the velocity of a reaching action depending on what they intend to do with the object once it is grasped. Despite these findings and irrespective of the larger context in which the action occurs, current models (e.g., Fitts’ law) target single, isolated actions, as, for example, pointing to a goal. In the current paper, we develop and empirically test a more ecologically valid model of two-step action planning. More specifically, 61 18-month olds took part in a reach-to-place task and their reaching and placing durations were measured with a motion-capture system. Our model explained the highest amount of variance in placing duration and outperformed six previously suggested models, when using model comparison. We show that including parameters of the first action step, here the duration of the reaching action, can improve the description of the second action step, here the duration of the placing action. This move towards more ecologically valid models of action planning contributes knowledge as well as a framework for assessing human machine interactions. The TWAIN model provides an updated way to quantify motor learning by the time these abilities develop, which might help to assess performance in typically developing human children.

Regulation of phosphate uptake kinetics in the bloom-forming dinoflagellates prorocentrum donghaiense with emphasis on two-stage dynamic process

Phosphorus is an essential element for the growth and reproduction of algae. In recent years, the frequent outbreaks of algal blooms caused by eutrophication have drawn much attention to the influence of phosphate (P) uptake on the growth of algal cells. The previous study only considered the effect of total P pools on the P uptake process of algal cells and considered the process as one stage, which is insufficient. P uptake by algae is actually a two-stage kinetic process because in many algae species, surface-adsorbed P pools account for a large proportion of total P pools. In this paper, we fit one-stage and two-stage models of P uptake by algae to our experimental data on short-term uptake kinetics of algae Prorocentrum donghaiense under P-deplete and P-replete conditions at 24°C. According to the experimental results, P. donghaiense possesses different P uptake characteristics under different P concentrations. P. donghaiense grows faster and exponentially for longer periods of time under P-replete condition. Ranges of change of Q_c (cell quota of intracellular P) and S_p (cell quota of surface-adsorbed P) during the culture time are obviously larger under P-replete condition than those under P-deplete condition. The value of K_p (represents the impact of P-starvation on P uptake rate) in one-stage model under P-deplete condition is smaller than that under P-replete condition, which is opposite to results of two-stage model and does not meet the actual biological significance of K_p. The two-stage model gives more reasonable and realistic explanations to the process of P uptake by algae no matter from the perspective of intuitive fitting effect, biological significance of parameters, statistical test results or essential dynamic process. These results, combined with long-term lab and field data in ocean, could be used to effectively predict algal blooms.

Development of modified HCH-1 kinetic model for long-term enzymatic cellulose hydrolysis and comparison with literature models

BACKGROUND

Enzymatic hydrolysis is a major step for cellulosic ethanol production. A thorough understanding of enzymatic hydrolysis is necessary to help design optimal conditions and economical systems. The original HCH-1 (Holtzapple-Caram-Humphrey-1) model is a generalized mechanistic model for enzymatic cellulose hydrolysis, but was previously applied only to the initial rates. In this study, the original HCH-1 model was modified to describe integrated enzymatic cellulose hydrolysis. The relationships between parameters in the HCH-1 model and substrate conversion were investigated. Literature models for long-term (> 48 h) enzymatic hydrolysis were summarized and compared to the modified HCH-1 model.

RESULTS

A modified HCH-1 model was developed for long-term (> 48 h) enzymatic cellulose hydrolysis. This modified HCH-1 model includes the following additional considerations: (1) relationships between coefficients and substrate conversion, and (2) enzyme stability. Parameter estimation was performed with 10-day experimental data using α-cellulose as substrate. The developed model satisfactorily describes integrated cellulose hydrolysis data taken with various reaction conditions (initial substrate concentration, initial product concentration, enzyme loading, time). Mechanistic (and semi-mechanistic) literature models for long-term enzymatic hydrolysis were compared with the modified HCH-1 model and evaluated by the corrected version of the Akaike information criterion. Comparison results show that the modified HCH-1 model provides the best fit for enzymatic cellulose hydrolysis.

CONCLUSIONS

The HCH-1 model was modified to extend its application to integrated enzymatic hydrolysis; it performed well when predicting 10-day cellulose hydrolysis at various experimental conditions. Comparison with the literature models showed that the modified HCH-1 model provided the best fit.