Emotion Detection Classifiers from EEG Signals using Support Vector Machines (SVM) and k-Nearest Neighbor (k-NN)

Authors

  • Anju Das , Neelima N

Abstract

—Information about the passionate condition of clients was becoming more and more relevant in
human-machine cooperation and brain-computer interface. The Mental health issue is growing rapidly in
these recent years. Adolescents and youthful grown-ups matured 16–30 years of age are the most highly
known casualties. Psychological well-being is an extremely difficult issue concerning passionate wellbeing.
This paper examines EEG-based Emotion Recognition using Support Vector Machines (SVM) and kNearest neighbor (k) calculation as classifiers. Evaluations are carried out to acquire the EEG signals of the
topic characterizing mental conditions, happiness, calm, sadness and anxiety. After the EEG signals are preprocessed, various types of EEG features are analyzed to build an emotion detection system. Results of the
investigation indicate that an average test accuracy of 66.51 per cent can be obtained by using four
emotional states for the classification Support vector machines and k-NN with precision of 52 percent

Published

2020-04-29

Issue

Section

Articles