TU Wien:Machine Learning VU (Musliu)/Exam 2024-01-23

Part 1

True or False single choice questions (mostly already available in older exams) +2 points for correct answer, 0 points for no answer, -1 point for incorrect answer

Part 2

very simple calculations (no calculator needed) for topics such as k armed bandits, root means squared error, 1R feature selection and convolution with stride 2 +3 points for correct answer, 0 points for no answer, -1.5 points for incorrect answer

Part 3

multiple choice questions on topics like feature selection, image augmentation or preprocessing. one question was what to do with missing values in test and training set. ? points for correct answer, 0 points for incorrect OR incomplete answer

Is Gradient Boosting 0 Rule?

Something about Monte Carlo Reenforcement Learning

What is Data Augmentation?

Adaboost weights initialization uniform