Diagnostic Tool for mental health

Inputs

1. EEG signals
2. Peripheral physiological signals
3. Battery test

Outputs

1. Emotional Profiling of the subject
2. Recommendation

Task Assigned

1. Learn about EEG related characteristic from other EEG data
   • Arithematic tasks
   • Visual Tasks
   • Motor tasks
2. Learned machine will then be used for Emotional profiling.
3. Hence, fine tuned machine by using transfer learning will be used to give muli-labelling output.

Tasks

Week1:

1. Understood the project and learned basics of machine learning and what transfer learning is.
2. Understood what MNIST database is and how to implement it.
3. Implemented MNIST database using KNN.
4. Changed parameters of the model to analyise the change in output

Week2:

1. Explored more about EEG related data/signals are classified and dealt with.
2. Read the documentation of motor tasks dataset and tried to analysis and how to deal with the signals.

Week3:

1. Loading and preprocessing of data from mne library

fnames = eegbci.load_data(subject, runs)

Week4:

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.

References

1. For understanding dataset: official documentation
2. Understanding multi label classification: Here