Response threshold variance as a basis of collective rationality

Collective decision-making during reproduction in social insects: a conceptual model for queen supersedure in honey bees (Apis mellifera)

Insect societies have served as excellent examples for co-ordinated decision-making. The production of sexuals is the most important group decision that social insects face since it affects both direct and indirect fitness. The behavioral processes by which queens are selected have been of particular interest since they are the primary egg layers that enable colony function. As a model system, previous research on honey bee reproduction has focused on swarming behavior and nest site selection. One significant gap in our knowledge of the collective decision-making process over reproduction is how daughter queens simply replace old or failing queens (=supersedure) rather than being reared for the purposes of colony fission (=swarming) or queen loss (=emergency queen rearing). Here, I present a conceptual model that provides a framework for understanding the collective decisions by colonies to supersede their mother queens, as well as provide some key recommendations on future empirical work.

Heterogeneous Thresholds, Social Ranking, and the Emergence of Vague Categories

Threshold models in which an individual's response to a particular state of the world depends on whether an associated measured value exceeds a given threshold are common in a variety of social learning and collective decision-making scenarios in both natural and artificial systems. If thresholds are heterogeneous across a population of agents, then graded population level responses can emerge in a context in which individual responses are discrete and limited. In this article, I propose a threshold-based model for social learning of shared quality categories. This is then combined with the voting model of fuzzy categories to allow individuals to learn membership functions from their peers, which can then be used for decision-making, including ranking a set of available options. I use agent-based simulation experiments to investigate variants of this model and compare them to an individual learning benchmark when applied to the ranking problem. These results show that a threshold-based approach combined with category-based voting across a social network provides an effective social mechanism for ranking that exploits emergent vagueness.

Inhibitory signaling in collective social insect networks, is it indeed uncommon?

Individual entities across levels of biological organization interact to reach collective decisions. In centralized neuronal networks, competing neural populations commonly accumulate information over time while increasing their own activity, and cross-inhibiting other populations until one group passes a given threshold. In social insects, there is good evidence for decisions mediated by positive feedbacks, but we found evidence for similar inhibitory signals only in honey bee (Apis mellifera) stop signals, and Pharaoh's ant- (Monomorium pharaonic) repellent pheromones, with only the former occasionally being used as cross-inhibition. We discuss whether these differences stem from insufficient research effort or represent genuine differences across levels of biological organization.

MRI-based machine learning model: A potential modality for predicting cognitive dysfunction in patients with type 2 diabetes mellitus

Background: Type 2 diabetes mellitus (T2DM) is a crucial risk factor for cognitive impairment. Accurate assessment of patients' cognitive function and early intervention is helpful to improve patient's quality of life. At present, neuropsychiatric screening tests is often used to perform this task in clinical practice. However, it may have poor repeatability. Moreover, several studies revealed that machine learning (ML) models can effectively assess cognitive impairment in Alzheimer's disease (AD) patients. We investigated whether we could develop an MRI-based ML model to evaluate the cognitive state of patients with T2DM. Objective: To propose MRI-based ML models and assess their performance to predict cognitive dysfunction in patients with type 2 diabetes mellitus (T2DM). Methods: Fluid Attenuated Inversion Recovery (FLAIR) of magnetic resonance images (MRI) were derived from 122 patients with T2DM. Cognitive function was assessed using the Chinese version of the Montréal Cognitive Assessment Scale-B (MoCA-B). Patients with T2DM were separated into the Dementia (DM) group (n = 40), MCI group (n = 52), and normal cognitive state (N) group (n = 30), according to the MoCA scores. Radiomics features were extracted from MR images with the Radcloud platform. The variance threshold, SelectKBest, and least absolute shrinkage and selection operator (LASSO) were used for the feature selection. Based on the selected features, the ML models were constructed with three classifiers, k-NearestNeighbor (KNN), Support Vector Machine (SVM), and Logistic Regression (LR), and the validation method was used to improve the effectiveness of the model. The area under the receiver operating characteristic curve (ROC) determined the appearance of the classification. The optimal classifier was determined by the principle of maximizing the Youden index. Results: 1,409 features were extracted and reduced to 13 features as the optimal discriminators to build the radiomics model. In the validation set, ROC curves revealed that the LR classifier had the best predictive performance, with an area under the curve (AUC) of 0.831 in DM, 0.883 in MIC, and 0.904 in the N group, compared with the SVM and KNN classifiers. Conclusion: MRI-based ML models have the potential to predict cognitive dysfunction in patients with T2DM. Compared with the SVM and KNN, the LR algorithm showed the best performance.

Adaptive phenotypic variation among clonal ant workers

Phenotypic variations are observed in most organisms, but their significance is not always known. The phenotypic variations observed in social insects are exceptions. Genetically based response threshold variances have been identified among workers and are thought to play several important adaptive roles in social life, e.g. allocating tasks among workers according to demand, promoting the sustainability of the colony and forming the basis of rationality in collective decision-making. Several parthenogenetic ants produce clonal workers and new queens by asexual reproduction. It is not clearly known whether such genetically equivalent workers show phenotypic variations. Here, we demonstrate that clonal workers of the parthenogenetic ant Strumigenys membranifera show large threshold variances among clonal workers. A multi-locus genetic marker confirmed that colony members are genetic clones, but they showed variations in their sucrose response thresholds. We examined the changing pattern of the thresholds over time generating hypotheses regarding the mechanism underlying the observed phenotypic variations. The results support the hypothesis that epigenetic modifications that occur after eclosion into the adult form are the cause of the phenotypic variations in this asexual species.

Nature of collective decision-making by simple yes/no decision units

The study of collective decision-making spans various fields such as brain and behavioural sciences, economics, management sciences, and artificial intelligence. Despite these interdisciplinary applications, little is known regarding how a group of simple ‘yes/no’ units, such as neurons in the brain, can select the best option among multiple options. One prerequisite for achieving such correct choices by the brain is correct evaluation of relative option quality, which enables a collective decision maker to efficiently choose the best option. Here, we applied a sensory discrimination mechanism using yes/no units with differential thresholds to a model for making a collective choice among multiple options. The performance corresponding to the correct choice was shown to be affected by various parameters. High performance can be achieved by tuning the threshold distribution with the options’ quality distribution. The number of yes/no units allocated to each option and its variability profoundly affects performance. When this variability is large, a quorum decision becomes superior to a majority decision under some conditions. The general features of this collective decision-making by a group of simple yes/no units revealed in this study suggest that this mechanism may be useful in applications across various fields.