Keep in mind that the coding is done to fulfill demands from professors, not necessarily to be elegant and clean :)
Project in TMA4315 Numerical mathematics. Given red/blue points with coordinates in R^4, make a neural network to predict the color of new points.
Project in TMA4268 Statistical learning. Fit different models to a dataset and evaluate the models. Fitted K-nearest neighbours, decision trees, support vector machine, neural network.
Project in TMA4180 Optimisation. Render points in R^2 and label them 0 or 1. Then use different optimisation algorithms to find an (or the) ellipse containing as many 1-points as possible and as few 0-points as possilble. Used gradient descent, conjugate gradient, bfgs and a barrier method.