Statistical Learning

Submitted by ruckdeschel on Mon, 01/21/2019 - 14:57
Person(s) in Charge
Prof.Dr. Peter Ruckdeschel & M.Sc. Tino Werner
Topic
Statistical Learning: Methods and Applications
Type
lecture (3) + exercise course (1)
KP
6KP
Workload
180h
Hours/Week
4.00h
Exam Type
extended exercise in R / homework project
Access Limits
None
Prerequisites
builds up on basic courses in Statistics/Stochastics
Link
for details, see StudIP at Oldenburg
Location
Uni Oldenburg, Mo, 10:00-12:00, Tu, 08:00-10:00 at W01 0-012
Description
  • Prediction
    •  linear and generalized linear models with more predictors than observations (lm, glm)
    •  regularized regression: enet, LASSO
    •  SVM regression
  • Classification
    • linear and quadratic discriminant analysis (LDA/QDA)
    • Support Vector Classification
    • CART
  • Resampling / Ensemble Methods
    • Bagging
    • Boosting
    • Random Forests
  • Outlook:  Ranking, Online Learning

All methods are introduced together with implementations in R and illustrated at applications

Semester
SoSe2019