Contains machine learning projects which were part of the course 'Introduction to Machine Learning' at ETH Zurich in spring semester 2021.
Authors: Robin Schmid, Marvin Harms, Pascal Müller
Simple linear regression as starter project.
Find optimal lambda for ridge regression, which is performed with a cross-validation.
Linear regressions (Lasso, Ridge, ElasticNet, Linear) using feature transformation.
Predict evolution of hospital patients and needs during their stay based on measurements in the first 12h.
Includes preprocessing of data (missing values, imbalance), binary classification evaluated with AUROC and regression using R2-score.
Classify protein mutations under abstract aspects, evaluation using F1-score.
Image taste classification using a convolution neural net.