Statistical Learning

Lecturer:

Imma Curato

Class Teacher:

Dirk Brandes

News:

The exercise class takes place every two weeks.

• Lecture: Monday, 16:00-18:00, He18 - 1.20
• First Lecture: 15/10/2018
• Exercise class: Thursday, 08:00-10:00, He18 - 1.20, biweekly
• First Exercise class: 08/11/2018

Final Exam:

oral exam of 20 minutes.

To participate in the oral exam, you have to register at campusonline.uni-ulm.de. 

Prerequisites:

Analysis I+II, Elementary Statistics and Probability, Stochastic I, and Measure Theory.

Learning Objective:

By attending the course you will 

• understand and master fundamental principles and modelling techniques for the analysis of regression and classification problems
• Gain or deepen, respectively, model assessment and inference techniques for linear and non-linear models.
• Exercising the acquired techniques by means of real data sets and the R software.

Contents:

This course covers topics of statistical learning in a mathematical and economical approach.

Specific topics are

• Linear Regression
• Classification
• Model assessment, selection and inference: cross-validation, bootstrap
• Regularization methods: Ridge and Lasso regression
• Overview of non-linear models: splines, support vector machines and neural networks

 Literature:

• T. Hastie, R. Tibshirani & J. Friedman, The Elements of Statistical Learning: data mining, inference and prediction, 2nd edition, Springer, 2009.
• G. James, D. Witten, T. Hastie & R. Tibshirani, An Introduction to Statistical Learning with Applications in R, Springer, 2013.
• W.H. Green, Econometric Analysis (Seventh Edition), Pearson, 2012.
• D.W. Hosmer, S. Lemeshow, R.X. Sturdivant, Applied Logistic Regression (Third Edition), 2013.
• G. Casella, R.L. Berger, Statistical Inference (Second Edition), 2001.
• B. Efron and R.J. Tibshirami, An Introduction to the Bootstrap, Chapman & HALL/CRC, 1994.

Exercise Sheets:

Moodle 

Moodle