Linear and nonlinear ARMA model parameter estimation using an artificial neural network

Learning spiking neuronal networks with artificial neural networks: neural oscillations

First-principles-based modelings have been extremely successful in providing crucial insights and predictions for complex biological functions and phenomena. However, they can be hard to build and expensive to simulate for complex living systems. On the other hand, modern data-driven methods thrive at modeling many types of high-dimensional and noisy data. Still, the training and interpretation of these data-driven models remain challenging. Here, we combine the two types of methods to model stochastic neuronal network oscillations. Specifically, we develop a class of artificial neural networks to provide faithful surrogates to the high-dimensional, nonlinear oscillatory dynamics produced by a spiking neuronal network model. Furthermore, when the training data set is enlarged within a range of parameter choices, the artificial neural networks become generalizable to these parameters, covering cases in distinctly different dynamical regimes. In all, our work opens a new avenue for modeling complex neuronal network dynamics with artificial neural networks.

Extended B-polynomial neural network for time-delayed system modeling using sampled data

 The combination of machine learning and artificial intelligent has already proved its potential in achieving remarkable results for modeling unknown systems. These techniques commonly use enough data samples to train and optimize their architectures. In the present era, with the availability of enough storage and computation power, the machine learning based data-driven system modeling approaches are getting popular as they do not interrupt the normal system operations and work solely on collected data. This work proposes a data-driven parametric neural network technique for modeling time-delayed systems, which is demanding but challenging area of research and comes under nonlinear optimization problem. The key contribution of this work is the inclusion of an extended B-polynomial into the network structure for estimating time-delayed first and second order system models. These type of models extensively used for addressing simulations, predictions, controlling and monitoring related issues. Also, an adaptive learning based convergence of the proposed algorithm is proved with the help of the Lyapunov stability theory. The proposed algorithm compared with existing techniques on some well-known example problems. A real practical system plant is also included for validating the proposed concept.

Post-Processing of High Formwork Monitoring Data Based on the Back Propagation Neural Networks Model and the Autoregressive—Moving-Average Model

Many high formwork systems are currently equipped with health monitoring systems, and the analysis of the data obtained can determine whether high formwork is a hazard. Therefore, the post-processing of monitoring data has become an issue of widespread concern. In this paper, we discussed the fitting effect of the symmetrical high formwork monitoring data using the autoregressive–moving-average (ARMA) model and the back propagation neural networks (BPNN) combined model to process. In the actual project, the symmetry of the high formwork system allows the analysis of local monitoring results to be well extended to the whole. For the establishment of the ARMA model, the accurate judgment of the model order has a significant impact. In this paper, back propagation neural networks (BPNN) are used to simulate the ARMA process. The order of the ARMA model is estimated by determining the optimal neural network structure, which is suitable for linear or nonlinear sequences. We validated this approach from the ARMA model data simulated in Monte Carlo and compared it with the Akaike information criterion (AIC) and Bayesian information criterion (BIC). The length of the sequence, the coefficients and the order of the ARMA model are considered as factors that influence the judgment effect. Under different conditions, the BPNN always shows an accuracy rate of more than 90%, while the BIC only has a higher accuracy rate when the model order is low and the judgment efficiency of the AIC is below 50%. Finally, the proposed method successfully modeled the stress sequence and obtained the stress change trend. Compared with AIC and BIC, the efficiency of the processing time series is increased by about 50% when an order is obtained by BPNN.

Arterial blood pressure feature estimation using photoplethysmography

Continuous and noninvasive monitoring of blood pressure has numerous clinical and fitness applications. Current methods of continuous measurement of blood pressure are either invasive and/or require expensive equipment. Therefore, we investigated a new method for the continuous estimation of two main features of blood pressure waveform: systolic and diastolic pressures. The estimates were obtained from a photoplethysmography signal as input to the fifth order autoregressive moving average models. The performance of the method was evaluated using beat-to-beat full-wave blood pressure measurements from 15 young subjects, with no known cardiovascular disorder, in supine position as they breathed normally and also while they performed a breath-hold maneuver. The level of error in the modeling and prediction estimates during normal breathing and breath-hold maneuvers, as measured by the root mean square of the residuals, were less than 5 mmHg and 11 mm Hg, respectively. The mean of model residuals both during normal breathing and breath-hold maneuvers was considered to be less than 3.2 mmHg. The dependency of the accuracy of the estimates on the subject data was assessed by comparing the modeling errors for the 15 subjects. Less than 1% of the models showed significant differences (p < 0.05) from the other models, which indicates a high level of consistency among the models.

NON-INVASIVE NOCTURNAL HYPOGLYCEMIA DETECTION FOR INSULIN-DEPENDENT DIABETES MELLITUS USING GENETIC FUZZY LOGIC METHOD

Hypoglycemia, or low blood glucose, is the most common complication experienced by Type 1 diabetes mellitus (T1DM) patients. It is dangerous and can result in unconsciousness, seizures and even death. The most common physiological parameter to be effected from hypoglycemic reaction are heart rate (HR) and correct QT interval (QTc) of the electrocardiogram (ECG) signal. Based on physiological parameters, a genetic algorithm based fuzzy reasoning model is developed to recognize the presence of hypoglycemia. To optimize the parameters of the fuzzy model in the membership functions and fuzzy rules, a genetic algorithm is used. A validation strategy based adjustable fitness is introduced in order to prevent the phenomenon of overtraining (overfitting). For this study, 15 children with 569 sampling data points with Type 1 diabetes volunteered for an overnight study. The effectiveness of the proposed algorithm is found to be satisfactory by giving better sensitivity and specificity compared with other existing methods for hypoglycemia detection.

Nonlinear dynamic system identification using Chebyshev functional link artificial neural networks

A computationally efficient artificial neural network (ANN) for the purpose of dynamic nonlinear system identification is proposed. The major drawback of feedforward neural networks, such as multilayer perceptrons (MLPs) trained with the backpropagation (BP) algorithm, is that they require a large amount of computation for learning. We propose a single-layer functional-link ANN (FLANN) in which the need for a hidden layer is eliminated by expanding the input pattern by Chebyshev polynomials. The novelty of this network is that it requires much less computation than that of a MLP. We have shown its effectiveness in the problem of nonlinear dynamic system identification. In the presence of additive Gaussian noise, the performance of the proposed network is found to be similar or superior to that of a MLP. A performance comparison in terms of computational complexity has also been carried out.

Natural occurrence of nocturnal hypoglycemia detection using hybrid particle swarm optimized fuzzy reasoning model

INTRODUCTION

Low blood glucose (hypoglycemia) is a common and serious side effect of insulin therapy in patients with diabetes. This paper will make a contribution to knowledge in the modeling and design of a non-invasive hypoglycemia monitor for patients with type 1 diabetes mellitus (T1DM) using a fuzzy-reasoning system.

METHODS

Based on the heart rate and the corrected QT interval of the electrocardiogram (ECG) signal, we have developed a hybrid particle-swarm-optimization-based fuzzy-reasoning model to recognize the presence of hypoglycemic episodes. To optimize the fuzzy rules and the fuzzy-membership functions, a hybrid particle-swarm-optimization with wavelet mutation operation is investigated.

RESULTS

From our clinical study of 16 children with T1DM, natural occurrence of nocturnal-hypoglycemic episodes was associated with increased heart rates and increased corrected QT intervals. All the data sets were collected from the Government of Western Australia's Department of Health. All data were organized randomly into a training set (8 patients with 320 data points) and a testing set (another 8 patients with 269 data points). To prevent the phenomenon of overtraining, we separated the training set into 2 sets (4 patients in each set) and a fitness function was introduced for this training process. The testing performances of the proposed algorithm for detection of advanced hypoglycemic episodes (sensitivity=85.71% and specificity=79.84%) and hypoglycemic episodes (sensitivity=80.00% and specificity=55.14%) were given.

CONCLUSION

We have investigated the detection for the natural occurrence of nocturnal hypoglycemic episodes in T1DM using a hybrid particle-swarm-optimization-based fuzzy-reasoning model with physiological parameters. In this study, no restricted environment (e.g. patient's dietary requirements) is required. Furthermore, the sampling time is between 5 and 10 min. To conclude, we have shown that the testing performances of the proposed algorithm for detection of advanced hypoglycemic and hypoglycemic episodes for T1DM patients are satisfactory.

Effects of simulated microgravity on closed-loop cardiovascular regulation and orthostatic intolerance: analysis by means of system identification

Microgravity-induced orthostatic intolerance (OI) continues to be a primary concern for the human space program. To test the hypothesis that exposure to simulated microgravity significantly alters autonomic nervous control and, thus, contributes to increased incidence of OI, we employed the cardiovascular system identification (CSI) technique to evaluate quantitatively parasympathetic and sympathetic regulation of heart rate (HR). The CSI method analyzes second-to-second fluctuations in noninvasively measured HR, arterial blood pressure, and instantaneous lung volume. The coupling mechanisms between these signals are characterized by using a closed-loop model. Parameters reflecting parasympathetic and sympathetic responsiveness with regard to HR regulation can be extracted from the identified coupling mechanisms. We analyzed data collected from 29 human subjects before and after 16 days of head-down-tilt bed rest (simulated microgravity). Statistical analyses showed that parasympathetic and sympathetic responsiveness was impaired by bed rest. A lower sympathetic responsiveness and a higher parasympathetic responsiveness measured before bed rest identified individuals at greater risk of OI before and after bed rest. We propose an algorithm to predict OI after bed rest from measures obtained before bed rest.

A new algorithm for linear and nonlinear ARMA model parameter estimation using affine geometry

A linear and nonlinear autoregressive (AR) moving average (MA) (ARMA) identification algorithm is developed for modeling time series data. The new algorithm is based on the concepts of affine geometry in which the salient feature of the algorithm is to remove the linearly dependent ARMA vectors from the pool of candidate ARMA vectors. For noiseless time series data with a priori incorrect model-order selection, computer simulations show that accurate linear and nonlinear ARMA model parameters can be obtained with the new algorithm. Many algorithms, including the fast orthogonal search (FOS) algorithm, are not able to obtain correct parameter estimates in every case, even with noiseless time series data, because their model-order search criteria are suboptimal. For data contaminated with noise, computer simulations show that the new algorithm performs better than the FOS algorithm for MA processes, and similarly to the FOS algorithm for ARMA processes. However, the computational time to obtain the parameter estimates with the new algorithm is faster than with FOS. Application of the new algorithm to experimentally obtained renal blood flow and pressure data show that the new algorithm is reliable in obtaining physiologically understandable transfer function relations between blood pressure and flow signals.