- EEG signals
- Peripheral physiological signals
- Battery test
- Emotional Profiling of the subject
- Recommendation
- Learn about EEG related characteristic from other EEG data
- Arithematic tasks
- Visual Tasks
- Motor tasks
- Learned machine will then be used for Emotional profiling.
- Hence, fine tuned machine by using transfer learning will be used to give muli-labelling output.
- Understood the project and learned basics of machine learning and what transfer learning is.
- Understood what MNIST database is and how to implement it.
- Implemented MNIST database using KNN.
- Changed parameters of the model to analyise the change in output
- Explored more about EEG related data/signals are classified and dealt with.
- Read the documentation of motor tasks dataset and tried to analysis and how to deal with the signals.
- Loading and preprocessing of data from mne library
fnames = eegbci.load_data(subject, runs)
1.Classified the files according to events and added them in a variable task
.
2. Prepared y matrix
3. Split the data in train and test sets.
4. Built a CNN model using Tensorflow and keras for multilabel classification.
- For understanding dataset: official documentation
- Understanding multi label classification: Here