Neural Networks in Forecasting Disease Dynamics

Хасанов, А. Г.; Шайбаков, Д. Г.; Жернаков, С. В.; Меньшиков, А. М.; Бадретдинова, Ф. Ф.; Суфияров, И. Ф.; Сагадатова, Ю Р

doi:10.24060/2076-3093-2020-10-3-198-204

Cited by 6 publications

(1 citation statement)

References 5 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…В задачах классификации для заданного множества входных значений ставится в соответствии метка класса из категориальной целевой переменной. Прогнозирование контролируемости БА относится именно к задачам классификации, поскольку целевые переменные прогноза являются по своей природе категориальнымистепень контроля над заболеванием [18,19].…”

unclassified

Prediction of uncontrolled course of bronchial asthma in children based on polymorphisms of genes of signaling molecules of the immune system and detoxification genes

Suprun

2022

Bûlletenʹ fiziologii i patologii dyhaniâ

View full text Add to dashboard Cite

Aim. To study the possibility of predicting the asthma control at various stages of the development of the disease, possibly on the basis of taking into account the genetic polymorphisms of Toll-like receptors, cytokines and detoxification system genes using the statistical method of learning neural networks.Materials and methods. We examined 167 children with bronchial asthma. The degree of asthma control was determined, the following mutations were detected: TLR2-Arg753Glu, TLR4-Asp299Gly, TLR4-Ghr399Ile, TLR9-T1237C, TLR9-A2848G; IL4-C589T, IL6- C174G, IL10-G1082A, IL10-C592A, IL10-C819T, IL12B-A1188C, TNFa-G308A; GSTM, GSTT, GSTM/GSTT, GSTP1 Ile105Val, GSTP1 Ala114Val, by PCR. The STATISTICA Automated Neural Networks package was used to model neural networks.Results. The model is based on the MLP (15-9-3) multilayer perceptron architecture with a layer of 15 input neurons (by the number of analyzed variables), a hidden intermediate layer of 9 neurons and an output layer of 3 neurons by the number of values of the classified variable (control). The training algorithm was chosen by BFGS as the most adequate to the classification task. The error function is traditionally chosen as the sum of squared deviations. The activation function of output neurons is Softmax. The activation function of the intermediate layer is hyperbolic. The volume of the training sample was 88 sets. The volume of samples for testing and quality control of the model was 36 sets. The resulting model was able to predict 79.01% of the correct values of the target variable (the degree of asthma control).Conclusion. The application of the developed program makes it possible to predict the possibility of uncontrolled or partially controlled asthma at any stage of the disease, including preclinical and pre-nosological for groups with a high risk of asthma. This allows you to individually adjust the measures of secondary and even primary prevention of asthma within the personalization of therapeutic approaches.

show abstract

unclassified

Prediction of uncontrolled course of bronchial asthma in children based on polymorphisms of genes of signaling molecules of the immune system and detoxification genes

Suprun

2022

Bûlletenʹ fiziologii i patologii dyhaniâ

View full text Add to dashboard Cite

show abstract

Using the neural network in clinical systems

Gusynina

Shornikova

2021

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

The article examines the identification of human bone fractures using convoluted neural networks. The method of recognition of photographs of patients is intended for automated systems of identification and video recording of images. Convolutional neural networks have a number of advantages, such as invariability when reducing or increasing image size, immunity to photo movements and deviations, changes in image perspective, and many other image errors. In addition, convolutional neural networks allow you to combine neurons at a local level in two dimensions, connect photographic elements in any place, and also reduce the total number of weights. The work describes a multi-layer convolutional network. The layers of which it consists are divided into two types: convolutional and sub-selective. Of interest is the use of the principle of weighting in the work. This principle allows you to reduce the number of characteristics of the neural network that can be trained. Network training is based on the rule of minimizing empirical error. This rule is based on the algorithm of inverse error propagation. This algorithm provides an instant calculation of the gradient of a complex function of several variables in case the function itself is predefined. Neural network training is based on probabilistic method. This method leads to more optimal results due to interference in the restructuring of network weights. The work confirms the axiomatics of the applied neural network, its architecture and its learning algorithm.

show abstract

Introduction of innovative technologies in friction units of heavy-duty tribosystems and monitoring of their condition

Колесников

Koropets

Avilov

2021

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

The article describes studies of heavy-duty metal-polymer tribosystems: wheel-brake pad and pyatnik-podpyatnik of rolling stock, as well as spline couplings of the MI-26 helicopter tail rotor transmission. Tests of the wheel - brake pad system were carried out on an inertial stand with two-way braking at loads and speeds close to real operating conditions. Methods for modifying polymers, fillers, and nanoscale additives have been developed for the Pyatnik - podpyatnik tribosystem of rolling stock. To increase the wear resistance of work surfaces two-layer carbon fibers were applied to the spline couplings. DLC- coatings. Bench tests of these coatings showed a 4.5-fold reduction in wear when testing full-scale slots with a load of 30,000 kg. H and the number of completed cycles in 1,000,000. Methods for monitoring spline couplings based on the analysis of the frequency spectrum of the acoustic-emission (AE) vibration signal generated during the operation of the friction unit are considered. The results of studying the working state of spline couplings obtained by vibration diagnostics in the acoustic frequency range are presented. The state estimation is based on both the characteristics of the time signal and the transformation of the signal in the frequency domain using modal decomposition of the signal using Hilbert-Huang transformations. It is shown that for the effective for monitoring heavy-duty tribosystems, it is advisable to use neural networks.

show abstract

Neural Networks in Forecasting Disease Dynamics

Cited by 6 publications

References 5 publications

Prediction of uncontrolled course of bronchial asthma in children based on polymorphisms of genes of signaling molecules of the immune system and detoxification genes

Prediction of uncontrolled course of bronchial asthma in children based on polymorphisms of genes of signaling molecules of the immune system and detoxification genes

Using the neural network in clinical systems

Introduction of innovative technologies in friction units of heavy-duty tribosystems and monitoring of their condition

Contact Info

Product

Resources

About