Emotion is an internal state of a human being that arises due to interpersonal events. It plays an important role in perception, attention, memory, and decision-making. Emotional state analysis can provide relevant information to diagnose various mental health disorders. Electrodermal activity (EDA) is a widely used physiological signal to assess emotional states due to its non-invasiveness. It is a measure of continuous variation in the electrical characteristics of the skin which reflects the sympathetic nervous system activity. EDA signals are non-linear, nonstationary, and multi-component in nature. This research aims to explore the applicability of EDA signals for the characterization of arousal and valence emotional states. Optimal decomposition algorithms are employed for EDA analysis and time, time-frequency features are used to characterize the emotional states. Machine learning algorithms are also explored to discriminate various emotional states. In this seminar, key results extracted using signal processing techniques and their classification performances would be presented.