Περιγραφή θέματος
Introduction
- Machine learning as a field, clinical bioinformatics, text classification, spam detection, mind reading, autonomous systems, supervised learning, unsupervised learning, reinforcement learning, the problem of induction, model space, inductive bias, classification versus regression.
Basic probability and estimation theory
Statistical hypothesis testing
Hypotheses in machine learning, stochasticity, proving an hypothesis, the null hypothesis, test statistics, standard single hypothesis testing, rejection regions, rejection procedure, statistical errors, statistical significance, p-values.
Naive Bayes classifier
By Tom M. Mitchell. Focus only on slides regarding Naive Bayes.
Decision trees
Lecture by Adele Cutler.
KNN
1- Nearest Neighbor, K-Nearest Neighbor (KNN), cross validation, distance-weighted KNN, locally weighted averaging, locally weighted regression, euclidean distance, weighted euclidean distance, information gain, memory based methods.
Performance metrics
Model selection and performance estimation
- Slides by Andrew W. Moore
Neural networks
Bayesian Networks and Causal Discovery
Recitations
Hypothesis testing