Methods for Statistical Model Fitting - January 2022

Length: 10 hours - 2 cfu

Abstract

The course provides fundamental understanding on statistical model fitting using Maximum Likelihood Estimation (MLE) and Maximum a Posteriori (MAP) paradigms. The main topics will be:

i) review on MLE and MAP;

ii) Linear and Multiple Linear Regression with MLE/MAP;

iii) Fisher's information matrix and Cramér–Rao bound;

iv) ROC curve analysis and its probabilistic interpretation;

v) Logistic Regression with MLE/MAP for classification;

vi) Survival analysis and Cox proportional hazard model.

Dates & Venue

Lecturer:

Dr. Massimo Rivolta - Dipartimento di Informatica

Assessor:

Dr. Massimo Rivolta - Dipartimento di Informatica