Homework 4:
Probabilistic machine learning to discretely classify neural data into three discrete positions in space.
- Rigorously deriving machine learning model parameters [3 probabilistic models: Gaussian Normal (class agnostic covariance), Gaussian Normal (class specific covariance), Poisson].
- Implementing said parameters using numpy and matplotlib to train our model based on given data, then classify test data. Gaussian Normal (class agnostic cov) achieved 96% accuracy.
Homework 6:
Implementing OLE, Wiener, and Kalman filters on neural data to continuously decode the position of the monkey's hand in space. Applications include creating continuous cursor control using a brain-computer-interface (BCI).