| Autor: |
Natalya Bodrina, Konstantin V. Sidorov |
| Rok vydání: |
2020 |
| Předmět: |
|
| Zdroj: |
Software Engineering Perspectives in Intelligent Systems ISBN: 9783030633189 |
| DOI: |
10.1007/978-3-030-63319-6_40 |
| Popis: |
The paper considers a dynamic model of emotional reactions that allows monitoring the change in the direction of testee’s emotional development during an audiovisual stimulation, which activates the auditory and visual analyzer of a person. This problem was solved by forming a new system of features that characterizes the morphology of attractors restored from the electroencephalogram (EEG) time series. There is a description of a bioengineering system for specific experiments and the composition of the generated multimodal database with the patterns of recorded signals. The paper discusses the issues of selecting parameter values (time delay, embedding dimension, time series length) that are necessary for the correct attractor reconstruction. Changes in human emotional reactions are assessed by fuzzy estimates of three characteristics of two-dimensional projections of an attractor reconstructed from EEG patterns. The results of monitoring emotions are presented in as a matrix of increment estimates of attractor characteristics determined by an additional index scale, which allows generating a numerical estimate for each fuzzy set element using the term index and a membership function. The created algorithm allows processing a sequence of EEG patterns recorded with a time shift and forming verbal conclusions about the direction of the emotional reaction development. Further research might involve the creation of a tool for the tasks of monitoring and correcting emotional reactions and cognitive activity of a person. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
