Statistical learning is concerned with model inference for pattern recognition, prediction and diagnosis, within a probabilistic and statistical framework.
In this course, students will learn
- to pose a supervised learning problem (classification and regression) by formulating it as a statistical criteria optimization problem,
- develop an appropriate learning algorithm
- and evaluate the resulting classification or regression function. The main supervised learning models and algorithms (e.g. perceptron, SVM/SVR, tree, ensemble methods) will be studied, along with a few generative approaches. A short introduction to unsupervised learning will also be given.