Objectives
Statistics is the essence behind data science. It is clearly essential to have a deep understanding of the theory and the methods. This is a prerequisite before following a machine learning course.
Syllabus
- Elements of decision theory: risk, loss, decision rules
- Optimal decisions, unbiasedness, equivariance, sufficient statistics
- Pointwise estimator: Z-estimator, M-estimator
- Asymptotical results: law of large numbers, central limit theorem, consistency, asymptotic normality
- Maximum likelihood, Fisher information, Kullback Leibler, asymptotic optimality
- Tests: definitions, the Neyman-Pearson lemma, Uniformly Most Powerful test, p-value, two-sided tests
- Bayesian framework
- Non parametric tests
Evaluation
- homework
- exam
Langage : English
- Teaching coordinator: Castellano Margherita
- Teaching coordinator: Latouche Pierre
- Teaching coordinator: Meynard Charles
- Teaching coordinator: Naulet Zacharie
- Teaching coordinator: Perrin Alexandre