Martin Mächler: Catalogue data in Autumn Semester 2016 |
Name | Prof. em. Dr. Martin Mächler |
Address | Seminar für Statistik (SfS) ETH Zürich, HG GO 14.2 Rämistrasse 101 8092 Zürich SWITZERLAND |
Telephone | +41 44 632 34 08 |
maechler@stat.math.ethz.ch | |
URL | http://stat.ethz.ch/~maechler |
Department | Mathematics |
Relationship | Retired Adjunct Professor |
Number | Title | ECTS | Hours | Lecturers | |
---|---|---|---|---|---|
401-5640-00L | ZüKoSt: Seminar on Applied Statistics | 0 credits | 1K | M. Kalisch, P. L. Bühlmann, R. Furrer, L. Held, T. Hothorn, M. H. Maathuis, M. Mächler, L. Meier, N. Meinshausen, M. Robinson, C. Strobl | |
Abstract | About 5 talks on applied statistics. | ||||
Objective | See how statistical methods are applied in practice. | ||||
Content | There will be about 5 talks on how statistical methods are applied in practice. | ||||
Prerequisites / Notice | This is no lecture. There is no exam and no credit points will be awarded. The current program can be found on the web: http://stat.ethz.ch/events/zukost Course language is English or German and may depend on the speaker. | ||||
401-6221-00L | Nonparametric Regression Special Students "University of Zurich (UZH)" in the Master Program in Biostatistics at UZH cannot register for this course unit electronically. Forward the lecturer's written permission to attend to the Registrar's Office. Alternatively, the lecturer may also send an email directly to Link. The Registrar's Office will then register you for the course. | 1 credit | 1G | M. Mächler | |
Abstract | This course focusses on nonparametric estimation of probability densities and regression functions. These recent methods allow modelling without restrictive assumptions such as 'linear function'. These smoothing methods require a weight function and a smoothing parameter. Focus is on one dimension, higher dimensions and samples of curves are treated briefly. Exercises at the computer. | ||||
Objective | Knowledge on estimation of probability densities and regression functions via various statistical methods. Understanding of the choice of weight function and of the smoothing parameter, also done automatically. Practical application on data sets at the computer. | ||||
401-6245-00L | Data Mining Special Students "University of Zurich (UZH)" in the Master Program in Biostatistics at UZH cannot register for this course unit electronically. Forward the lecturer's written permission to attend to the Registrar's Office. Alternatively, the lecturer may also send an email directly to Link. The Registrar's Office will then register you for the course. | 1 credit | 1G | M. Mächler | |
Abstract | Block course only on prediction problems, aka "supervised learning". Part 1, Classification: logistic regression, linear/quadratic discriminant analysis, Bayes classifier; additive and tree models; further flexible ("nonparametric") methods. Part 2, Flexible Prediction: additive models, MARS, Y-Transformation models (ACE,AVAS); Projection Pursuit Regression (PPR), neural nets. | ||||
Objective | |||||
Content | "Data Mining" is a large field from which in this block course, we only treat so called prediction problems, aka "supervised learning". Part 1, Classification, recalls logistic regression and linear / quadratic discriminant analysis (LDA/QDA) and extends these (in the framework of 'Bayes classifier") to (generalized) additive (GAM) and tree models (CART), and further mentions other flexible ("nonparametric") methods. Part 2, Flexible Prediction (of continuous or "class" response/target) contains additive models, MARS, Y-Transformation models (ACE, AVAS); Projection Pursuit Regression (PPR), neural nets. | ||||
Lecture notes | The block course is based on (German language) lecture notes. | ||||
Prerequisites / Notice | The exercises are done exlusively with the (free, open source) software "R" (http://www.r-project.org). A final exam will also happen at the computers, using R (and your brains!). |