Login |
 
 

Overview

Hours per week : 2

Credits : 4

Participants: Min: 2 Max: 16 Expected: 6

Type of course: Seminar  (Bachelor/Master)

Language: english

Content

Do you think you know how to mine data? Do you want to learn how to apply the latest techniques for data mining? Learning a catalog of techniques, algorithms and buzzwords really only gets to be interesting when you get to apply them to real-world datasets. This is your chance!


Whether it's predicting movie preferences or forecasting the use of a bike sharing system, the data requires non-trivial analysis. As Nobel Laureate Niels Bohr said, "Predictions are hard, especially about the future." The skills necessary to make those hard predictions are honed only with practice.

In this seminar, we will be taking real-world datasets from the Kaggle website and potentially other sources, and will apply the steps necessary to perform intelligent data analysis. The students will work together in teams of two persons. During the semester, students will meet with their seminar adviser every two weeks (10% of the grade), they will give two short status updates (10% of the grade per update), hold a final presentation (30% of the grade), and write a short paper detailing
the work performed (40% of the grade.)

Form of Exam

oral presentation (30 minutes) and written examination at the end of semester

Preliminaries

Basic data mining knowledge, or visiting our parallel Data Analysis course is mandatory.