Practice Course: ML-PR Pattern Analysis and Machine Intelligence
Angewandte Informatik
Veranstaltungs-Nr: ML-PR CP: 8 SWS: 4 PR
Lehrform: Praktikum
Unterrichtssprache: Englisch
Selbststudium: 6 CP Kontaktstudium: 2 CP
Please email your student ID : teaching@ccc.cs.uni-Frankfurt.de ( with mail subject: ML-PR ) indicating interest to enroll in class within first 2 weeks of semester (until October 27, 2017).
Schedule:
Please bring a laptop to each session (ideally with Linux or OSX). We will install the needed python suites and packages together in the practice sessions.
| Monday, 12.02.2018, 10:00 – 16:00 |
SR 9, RM 11-15 |
- 10:00 Lecture, [pdf] Intro
- 11:00 Lecture, [pdf] Version-control, Styleguide, Code documentation
- 12:00 Practice, [ipynb] Python for datascience (basics, numpy, pandas), requirements
- 14:00 Practice, [pdf] Polynomial regression, Overfitting, mlpr1718_projections ,
[ipynb] Classifiers using pandas and sklearn
|
| Tuesday, 13.02.2018, 10:00-12:00 |
SR 9, RM 11-15 |
- Lecture, Introduction to kaggle, Presentation of an implementation for the San Francisco Crime Challenge, 4_san_francisco_crime
|
| Tuesday, 13.02.2018, 12:00 – 16:00 |
307, RM 11-15 |
- Practice, Implementation of your solutions for the San Francisco Crime Challenge
|
| Wednesday, 14.02.2018, 10:00 – 17:00 |
SR 11, RM 11-15 |
|
| Thursday, 15.02.2018, 10:00 – 17:00 |
SR 9, RM 11-15 |
- 10:00 Practice, CNNs on a practical example (continue in MLP PyTorch notebook)
- 12:00 Lecture, Interactive visualizations, GPU acceleration [tar]TensorBoardExamples
- 13:00 Lecture, Unsupervised learning [pdf] UnsupervisedML
- 14:00 Practice, Autoencoder and Convolutional Autoencoders (add to PyTorch notebook)
|
| Friday, 16.02.2018, 10:00 – 14:00 |
SR 9, RM 11-15 |
- 10:00 Project pitches, discussions, selection
- 12:00 Lecture, State-of-the-art overview of cutting edge research in ML
|