Neural Network Models for the Analysis and Visualization of Latent Dependencies: Examples of Psycho Diagnostic Data Processing

Slavutskaya, Elena; Славутский, Л. А.; Nikolaev, Evgeni L.; Zakharova, A.N.

doi:10.1007/978-3-030-65857-1_7

Cited by 7 publications

(2 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The system analysis of the multidimensional random data internal connections is of interest for a wide range of tasks. This applies both to the technical tasks of analyzing information flows [1] and to the problems of psychology, sociology [2,3], where it is necessary to analyze the hidden connections between heterogeneous data [4]. If the multidimensional data have different numerical scales, then it is difficult to assess the relationships between them using traditional statistical methods, such as correlation and regression analysis [5].…”

Section: Introductionmentioning

confidence: 99%

Machine learning techniques combination for selective and hierarchical analysis of psycho diagnostic data

Slavutskaya

Grigoryeva²,

Grigoryeva

et al. 2023

International Conference on Digital Transformation: Informatics, Economics, and Education (DTIEE2023)

View full text Add to dashboard Cite

The research is devoted to the use of the decision tree method and artificial neural networks for the analysis of psycho diagnostic data. The possibility of applying these two data mining tools together to obtain additional information is explored. The analysis is carried out on the example of the psychological testing results of preschoolers' parents. Two widely known tests were used, which allow to assess the relationship between 27 psychological indicators having different numerical scales. The sample consisted of 496 respondents. It is shown that the combined use of the indicated processing techniques allows for the selective and hierarchical assessment of the interrelationships of multidimensional heterogeneous data. The proposed approach is available to a wide range of specialists from different fields of knowledge, since the tools used are included in widespread software.

show abstract

Section: Introductionmentioning

confidence: 99%

Machine learning techniques combination for selective and hierarchical analysis of psycho diagnostic data

Slavutskaya

Grigoryeva²,

Grigoryeva

et al. 2023

International Conference on Digital Transformation: Informatics, Economics, and Education (DTIEE2023)

View full text Add to dashboard Cite

show abstract

“…Искусственные нейронные сети (ИНС) [2,4,8], методы машинного обучения и искусственного интеллекта (ИИ) [18,24,26,32] всё более широко применяются для анализа случайных процессов и сигналов [14,34], в частности в интеллектуальной энергетике [16,17,[21][22][23]. Многослойный персептрон (multilayer perceptron, MLP) является основой нейросетевых алгоритмов и при обучении «с учителем» представляет собой простейшую нейронную сеть прямого распространения [3,6,7]. Последние десятилетия структура используемых ИНС все более усложняется [13,29,31], применяются сверточные [19,30], рекуррентные [12] сети и т.д.…”

unclassified

Neural Network Signal Processing: The Tasks Without “Deep Learning”

Славутский

Slavutskaya

2023

Vestn. Čuv. univ.

View full text Add to dashboard Cite

The purpose of the study is to evaluate the possibility of using artificial neural networks for electrical signals processing. Methods. The application of a multilayer perceptron for these purposes as the simplest neural feed forward network is considered. Its difference as the basis of neural network algorithms is that when analyzing dynamic processes, signal processing should be carried out in a “sliding time window”. Results. It is shown that neural network processing makes it possible to approximate the shape of the signal with high accuracy and determine its parameters in real time. Using the example of periodic signals and transients in electrical circuits, accuracy assessments are made and the features of neural network processing are analyzed. The necessary sizes of the training sample of signals and the level of errors that occur when testing a neural network are discussed. Estimates of the required signal sampling frequency, the “sliding window” duration and the variation range of signal parameters when creating a training sample are given. Conclusions. It is shown that the proposed approach does not require “deep learning” of neural networks with complex architecture. It enables to create a signals training sample based on simple analytical formulas and to control the neural network algorithm quality at intermediate stages of calculations.

show abstract