DM825 - Introduction to Machine Learning

This is a naive demonstration for the k-fold cross validation. The k rectangles in the plot denote the k folds of data. Each time a fold will be used as the test set and the rest parts as the training set.

Schedule

Spring 2013, third quarter, weeks 5-11Monday 08:00-10:00 in IMADA Seminarrum
First lecture: January 28, 2013.Wednesday 16:00-18:00 in U49
Last lecture: March 15, 2013.Friday 08:00-10:00 in IMADA Seminarrum

Lectures

(see also syllabus below for a detailed list of contents)

Lec.DateTopicLiterature and Assignments
L004.11.2012Presentation
L128.01.2013Introduction, linear regression, k-nearest neighbor[B2 1-2.4; B10; B6 5.1-5.10; A0] [ exercises ]
L230.01.2013Linear models and probability interpretation[B1 sc1.1-1.4, 2.3.6, 3.1, 3.3; B10] [ exercises ]
L301.02.2013Binary variables, logistic regression, model assessment[B2 sc 2.1; B10; B2 sc 1.5.5, 3.2] [ exercises ]
E104.02.2013Exercises: solutions practical part, solutions theoretical part[B2 sc. 2.2, 3.3]
L406.02.2013Empirical model assessment, Generalized linear models[B2 sc7.1-7.3, 7.10-7.11; B1 sc2.4; B10] [ exer. / sol ]
L508.02.2013Perceptron, multi-layer perceptron, neural networks[B1 sc1.5, 4.0, 4.1.1, 4.1.7, 5.1, 5.2, 5.3] [ exercises ]
L611.02.2013Neural networks[B1 sc5.2, 5.3, 5.5; B2 ch11] [ exercises / sol ]
E213.02.2013Exercises[ Assignment 1 ]
L715.02.2013Gaussian discriminant analysis, naive Bayes[B10; B1 sc4.2; B2 4.1-4.3] [ exercises / sol ]
L818.02.2013Support vector machines[B10; B1 sc7.1, AppE; B2 4.5; A3] [ exercises / sol ]
L920.02.2013SVM and Kernel Methods[B10; B1 6.1, 6.2; B2 sc2.8.2, ch6, sc12.1-12.3.5; B11]
E322.02.2013Exercises[ exercises / sol ]
L1025.02.2013Application Guidelines, Feature Selection, Bagging, Boosting[B10, B12, B1 sc 14.1, 14.3]
L1127.02.2013Probabilistic graphical models[B1 sc 8.1; B13] [ exercises / sol ] [ Assignment 2 / sol ]
L1201.03.2013Probabilistic graphical models[B1 ch 8; L11] [ exercises / sol ]
E404.03.2013Exercises
L1306.03.2013Mixtures models, k-means, EM algorithm[B10; B1 sc 9.1, 9.2; L10] [ exercises / sol ]
E508.03.2013Exercises
L1411.03.2013Tree based methods, Principal component analysis[B1, sc 14.4, 12.1; B10; B2, sc 9.2, 14.5.1]
E613.03.2013Exercises[ exercises / sol ]
E715.03.2013Exercises
E803.04.2013Question time at 9:00 in U49D

Literature

Text book

Other references

Assessment

  • Mandatory assignments, pass/fail, internal evaluation by the teacher. The mandatory assignments include programming work. The assignments must be passed before the written exam can be attended.
  • Written exam: April 4, 2013, 10-13 in U49, U49B
    Exam instructions, Answer Templates [ Latex | Word | OpenDocument ]
    Solutions
  • Reexam: June 12, 2013
  • Syllabus
  • Exam 2010

Author: Marco Chiarandini <marco@imada.sdu.dk>

Date: 2013-05-01 17:30:57 CEST

HTML generated by org-mode 6.33x in emacs 23