Statistics
Objectives
The aim of this course is to enable the future engineer to build a mathematical model based on the observation of a random phenomenon and a collection of experimental or sampling data. This construction goes from the choice of model to its precise adjustment and its validation. This model must then allow a better understanding or analysis of the phenomenon and lead, if necessary, to decision-making or forecasts.
Description
1st course: Normal distributions; mean and variance.
2nd course: Basis of estimation; unbiased estimation of minimum variance.
3rd course: Fisher Information; Cramer-Rao inequality; maximum likelihood.
4th course: Basis of hypothesis testing; Lemma of Neyman and Pearson.
5th course: Likelihood ratio test; Linear regression. 6th course: Multilinear regression.
7th course: Complements; revisions. Tutorials
TD 1: Mean-variance independence, Gaussian case.
TD 2: Estimate (1).
TD 3: Estimation (2).
TD 4: Hypothesis testing.
Targeted skills
The random models are introduced in application contexts that are related to quality control, reliability, sample surveys or the monitoring of an industrial production or economic factor. The engineer who will have to adress such problems can validly adapt these models to his specific problem and solve it.
Bibliography
Garel, B. (2002) : Modélisation probabiliste et statistique. Cépadues Editions.
Garel, B. (2018) : Modèles mathématiques du hasard. Ellipses.
Lejeune, M. (2010) : Statistique, la théorie et ses applications. Springer.
Saporta, G. (2011) : Probabilités, analyse des données et Statistique. Editions Technip.
Session 1 ou session unique - Contrôle des connaissances
Modalité | Nature | Coefficient | Remarques |
---|---|---|---|
CT (contrôle terminal) | Oral/Ecrit | 70% | Statistiques - Examen |
CC (contrôle continu) | Rapport | 30% | Statistiques - BE |
Session 2 - Contrôle des connaissances
Modalité | Nature | Coefficient | Remarques |
---|---|---|---|
CT (contrôle terminal) | Oral/Ecrit | 70% | Statistiques - Examen |
CC (contrôle continu) | Rapport | 30% | Statistiques - BE |
Contact(s)
CHABERT MARIEPlaces
- Toulouse