Galvanic Skin Conductance Response and Bio Inspired Algorithms for Human Emotion Classification: A Study

Exploring and implying supervised machine learning algorithms inspired with the nature acts as the significant source of motivation. Exploring various natured inspired algorithms to optimize the results of the other algorithm as well is also one of the motivations. Due to their exploration and exploitation capabilities, bio-inspired algorithms are able to bypass the local optimum and locate the global optimum. The algorithms used to create emotion detection models at the present scenario is based on support vector machines, decision trees, random forests and neural networks. It is challenging task to achieve a breakthrough in classification accuracy as traditional emotion recognition does not fully reflect the hidden relationship between EEG signals and emotions states. The more abstract, deeper and discriminating relationship between the EEG states cannot be obtained only by the isomorphic data analysis. A psychological emotion recognition based on skin electric signal features and data was developed to address the aforementioned issues. Human skin is considered for discussion as it can be easily examined without expensive devices.