An Analytical Study on Computational Intelligence Algorithms for Determining Activation Levels in Typical Behaviors of The Autism
Machine learning algorithm, Performance analysis, Afective state, Stereotyped behavior, Autism
Stereotyped Behaviors (SBs) are atypical and repetitive movements of the body, which can be related to low mental health condition. These behaviors can lead the patients to increase their activation levels. The development of systems able to both recognize SBs and inferring activation level automatically can aid some therapeutic approaches. In this paper, a system is proposed to infer activation levels from recognized SBs, where different Machine Learning Algorithms (MLAs) are used for identifying the SBs and for determining the related activation levels. A performance metric, called Temporal Performance Index (TPI), is also proposed to evaluate the performance of MLAs that
consider the time for classification of SBs by relating it to accuracy and precision. For classifying the SB, the Hidden Markov Models and Multilayer Perceptron presented the best performance than Support Vector Machine and Convolutional Neural Network. The Adaptive Neuro-Fuzzy technique based on the Fuzzy C-Means Clustering algorithm allowed one to determine and differentiate the activation levels of the three stereotyped behaviors considered in the present
study.