This project applies different kinds of ML algorithms on the SEED Dataset to classify the data into 3 states: Positive Emotion, Neutral, Negative Emotion.
The algorithms in this repo are applied to the 5th dimension (gamma band which is best for emotion recognition) of “de_LDS” features of the Dataset. There are 62 inputs (feature vectors) in each de_LDS for every subject of the experiment. All de_LDS features contribute to more than 150000 samples in the dataset. The objective is to train models to map them using 62 feature vectors to 3 labels.
The project is written in python using anaconda in jupyter notebook. Some of the libraries youl'll need in the project are: Tensorflow, sk-Learn, Numpy, Pandas, Matplotlib and Seaborn.
Rasoul Ghaznavi
This project is licensed under the GNU General Public License v3.0.