Pattern recognition, machine learning
3 ECTS
Aim:This course introduces the fundamentals of statistical pattern recognition and machine learning. We first explain Bayesian decision theory, and show how a given PR problem may be expressed in terms of probabilities and distributions. We then study various well-known techniques usable to solve the problem raised as the outcome of formalization. Generalization theory, and elements of machine learning are introduced at last. Examples are provided on real problems, most of which arise from medical imaging or sensoring.
Content:
- Bayesian decision theory
- Maximum-likelihood based methods (naïve Bayes, EM, HMM)
- Linear discriminant analysis
- Multilayer neural networks
- Introduction to Support Vector Machines