Adrian Ulges

Machine Learning (WS 16/17)

News

• The final project presentations are now available.
• Some remarks regarding the final presentations
  • Plan for talk time (25 min) + live demo (5-10 min) + questions.
  • Check your laptop/projector beforehand (our schedule is tight).
• The time table for the final presentations on Feb 24 (Room C-407)
  • 10:00 - Sergej Bomke / Dennis Müller (Gaze Tracking)
  • 10:40 - Alexander Rubinov / Stephan Sigel (Gaze Tracking)
  • 11:20 - Alexander Krenz / Nadja Kurz (Gaze Tracking)
  • -- Lunch Break --
  • 13:00 - Christian Caspers / Felix Hamann (Wikification)
  • 13:40 - Thipee Thirunavukkarasu / Dennis Bingenheimer (Wikification)
  • 14:20 - Sven Bodemer / Alexander Haas (Wikification)
  • 15:00 - Martin Zakarian Khengi / Tiras Zemicael (Wikification)
• See these slides for more information on the course.

Slides

• Chapter 01: Introduction (scan with notes)
• Chapter 02: Decision Trees (scan with notes)
• Chapter 03: Features (scan with notes)
• Chapter 04: Logistic Regression (scan with notes)
• Chapter 05: Clustering (scan with notes)
• Chapter 06: Dimensionality Reduction (scan with notes)
• Chapter 07: Neural Networks I (scan with notes)
• Chapter 08: Neural Networks II (scan with notes)
• Chapter 09: Deep Learning (scan with notes)
• Chapter 10: Instance-based Learning (scan with notes)
• Chapter 11: Naive Bayes and Graphical Models (scan with notes)
• Chapter 12: Recommender Systems (scan with notes)

Exercises

• Exercise 01: sheet
• Exercise 01: classifier_simple.py
• Exercise 02: sheet
• Exercise 02: nytimes_classifier.zip
• Exercise 03: sheet
• Exercise 03: EM.zip
• Exercise 04: sheet
• Exercise 04: mnist.py
• Exercise 05: sheet
• Exercise 05: Caltech Faces Ground Truth
• Exercise 05: Sample Segmentations over Time (masks.zip)
• Exercise 05: sample solution for CNN face segmentation
• Exercise 06: sheet

Videos

... can be found in AMIGO (Login with HDS Account).

Other Materials

Python Resources

• Some nice slides on Python. Thanks to Prof. Dr. Peter Barth (MI) for sharing!
• A YouTube video Python for Java developers (nice and short).
• The official tutorial (a bit longer).

• For Numpy: The Quickstart Tutorial.
• For Pandas: The 10 Minutes to Pandas.

Schedule

Weeks 01-07

• 11:45-13:15 - Lecture (C-037)
• 14:15-15:45 - Lecture (C-035)
• 16:00-17:30 - Lab (C-213)

• 14:15-17:30 - Project Lab (C-213)

Literature

The main book for this course will be Stephen Marsland: Machine Learning - An Algorithmic Perspective (available in the library).

Kontakt

Adrian Ulges
Unter den Eichen 5, Haus C links, C-011
e-mail: adrian dot ulges at hs-rm dot de
Consultation: Mondays, 16:00 - 17:00