Classification of EEG signals using machine learning

This project involves the normalization of a dataset and feature extraction from preprocessed EEG data, followed by evaluation and optimization using a Support Vector Machine (SVM) classifier. The tasks include:

1.Feature Extraction and Normalization:

•Extract features from EEG data, such as:

a) Power Spectral Density (PSD)

b) Wavelet Transform

c) Differential Entropy

d) Hiorth Parameters (activity, mobility, complexity)

e) Additional statistics if required based on results

•Normalize the extracted features.

2.New Labeling System:

•Implement a new labeling system for a preprocessed dataset provided by the recruiter.

•Determine the most suitable loss function for handling an unbalanced dataset.

3.Comparative Studies:

•Combine two of the best feature types from the first task with two types of pre-trained CNN models.

•Conduct four experiments to compare results.