Machine Learning and Data Mining
[go to overview]Winter Terms 2016 / 2017
Basic Course Information
Lecture - Machine Learning and Data Mining
Course number: 0432028
| Lecturers: | Prof. Dr. Markus Strohmaier Dr. Florian Lemmerich Dr. Philipp Singer |
| Dates: | Wed 08.30-10.00 |
| Room: | KO Gebäude G - G 410 Room changed! |
Exercise - Machine Learning and Data Mining
Course number: 0432028
| Dozent(in) | Prof. Dr. Markus Strohmaier Dr. Philipp Singer Dr. Florian Lemmerich Evgeniy Vasilev |
| Dates: | Wed 10.15-11.45 |
| Room: | KO Gebäude G - G 410 Room changed! |
Lecture schedule:
This schedule might be subject to changes. Materials are for use in this course only and may not be redistributed.
|
Lecture # |
Lecture Topic |
Date |
Materials |
Comment |
|
1 |
Overview and Motivation |
2016-10-26 |
|
|
|
2 |
Classification (Task, Evaluation, Nearest neighbor) |
2016-11-02 |
|
|
|
3 |
Classification (Naive Bayes) |
2016-11-09 |
|
|
|
4 |
Classification (Decision Trees) |
2016-11-16 |
|
|
|
5 |
Classification (Ensemble Learning, SVM) |
2016-11-23 |
|
|
|
6 |
Clustering (Task, k-means, k-medoids) |
2016-12-07 |
|
|
|
7 |
Clustering (EM-Algorithm, density-based clustering) |
2016-12-14 |
|
|
|
8 |
Clustering (hierarchical clustering, other techniques) |
2016-12-14 |
|
|
|
9 |
Pattern Mining (Association Rules) |
2017-01-11 |
|
|
|
10 |
Pattern Mining (Subgroup Discovery) |
2017-01-18 |
|
|
|
11 |
Preprocessing / Matrix Factorization (PCA - SVD - NMF) |
2017-01-18 |
|
|
|
12 |
Sequential Data |
2017-02-01 |
|
|
|
13 |
Neural Networks and Deep Learning |
2017-02-08 |
|
|
|
14 |
Summary and Outlook |
2017-02-15 |
|
There will be no lecture at the 30th of November 2016. We will give the omitted lecture in the exercise slot on the 14th of December, instead.
Exercises:
|
Exercise # |
Topic | Material |
| 1 | Introduction to Python | notebook |
| 2 | Introduction to scikit-learn | notebook 1 notebook 2 |
| 3 | Naive Bayes | notebook |
| 4 | Decision Tree Feature Standardization pen & paper exercise |
notebook 4a notebook 4b exercise exercise_solution |
| 5 | Deep Learning | notebook_5 dataset |
Home assignments:
Assignments should be performed in the jupyter notebook, saved and sent to vasilev@uni-koblenz.de, florian.lemmerich@gesis.org, AND philipp.singer@gesis.org with Subject: [ML-Assignment]. You create groups on your own, names of all group participants should be mentioned in the letter. Group size: 2-3 persons. Deadlines are strict!
|
Assignment # |
Type | Deadline | Task, material | Example Solution |
| 1 | programming | Nov 8, 12:00h (noon) | notebook & dataset | solution |
| 2 | pen & paper | Nov 22, 23:59h (extended by 12 hours) |
task description | solution |
| 3 | programming | Nov 29, 12:00h (noon) | notebook | solution |
| G | graded programming | Jan 15, 23.59h | task & dataset | solution |
| 4 | pen &paper | Feb 7, 23.59h | task description | solution |
There will be interview talks for the graded home assignment on 25th of January in room A221 (office of Prof. Strohmaier).
You can find the exact schedule for the interview talks here.
Exam:
We hold the exam on February, 22nd at 08.30h in KO Gebäude E - E 011. Please be on time and bring a student id. You are allowed to bring a non-programmable calculator with you, but no other material.
Newsgroup:
You can use our newsgroup to ask any question about course content or course organization. Please also try and help out your fellow students with their issues by answering directly.
https://deliver.uni-koblenz.de/webnews/newsgroups.php?group=infko.mldm
Lecturers
- Alumnus