Course: Statistical Learning Methods

Prof. Jacques Savoy
University of Neuchatel
Computer Science Department

Main objective

The main objective of this course is to introduce the students to the various techniques and strategies that can be used to

to be able to use and program in R, the statistical language (used to analyze Big Data)
to select, apply and evaluate a learning method using R and to interpret the results
to select the most appropriate method according to the data, and to evaluate the quality of the fit.

Practical exercises will complete the theoretical presentation.

Description

Introduction to R (statistical software); statistical models and evaluation with R; regression simple and multiple; logistic regression; k-nearest neighbors; linear discriminant analysis (LDA); model evaluation, variable selection and regularization; resampling approaches & evaluation; support vector machines (SVM); boosting; unsupervised approaches (principal component analysis, multidimensional scaling).

The final mark is based on both a final written exam and the results of the practical exercices.

References

G. Jones, D. Witten, T. Hasti, R. Tibshirani: An Introduction to Statistical Learning. With Applications in R. Springer, 2013.
Trevor Hastie, Robert Tibshirani, Jerome Friedman: The Elements of Statistical Learning: Data Mining, Inference, and Prediction. Springer-Verlag, New York (NY), 2nd Ed., 2009.
Christopher M. Bishop: Pattern Recognition and Machine Learning. Springer, 2006.
M.J. Crawley: The R Book. 2nd Ed., Wiley, 2012

Additional information

The Machine Learning web site
The ML data sets web site (around 160 different data sets)
The random.org generation of values
The CA (correspondance analysis) package

To obtain different tools

The Weka software
The official R web site
The CRAN (Comprehensive R Archive Network) web site

Prof. Jacques Savoy
Université de Neuchâtel
Computer Science Department
Rue Emile-Argand 11
CH-2009 Neuchâtel
Switzerland

+41 32 718 1375 (phone)
+41 32 718 2701 (fax)