INF8245E - Machine Learning

Program overview Schedule Intro to ML, Linear Regression Linear Regression, Overfitting, ML Pipeline, Classification, k-NN, More on Regression,Gradient Descent Gradient Descent, Regularization, Decision theory, Empirical Risk Minimization Bias-Variance Tradeoff, linear classification GLM, GDA, Naive Bayes Logistic Regression, Newton-Raphson method, Evaluation Metrics, Perceptron Max-margin Classifiers, SVMs Decision trees, Ensembles: Bagging, Random Forests, Stackin Boosting, Neural Nets, Backpropagation Backpropagation, Deep Neural Networks, Convolutional Networks Optimization, Dimensionality Reduction, PCA, LDA Bayesian learning, MLE, MAP, Bayesian Linear Regressio Kernel Methods, Gaussian Process, Frontiers, What Next?

Probability and Statistics

Program overview Part I Axioms, conditional probability, Bayes rule, combinatorial analysis. Random variables: functions of distribution, mass and density, expectation and variance. Discrete and continuous probability laws. Reliability. Random vectors: correlation, multinormal distribution, central limit theorem Stochastic processes: Markov chains, Poisson process, Brownian motion. Descriptive statistics: diagrams, calculation of characteristics. Sampling distributions: estimate, mean squared error, confidence intervals. - Hypothesis tests: parametric tests and fit test.