Overview
The seminar will cover a variety of machine learning models and algorithms (including deep neural networks) and will discuss their applications in a diverse set of domains. Furthermore, the seminar will discuss how domain knowledge is integrated into vanilla ML models.
Announcements
- 22.09.2016
- Seminar Schedule goes online.
- 15.09.2016
- Seminar Website goes online.
- 15.09.2016
- Topics will be assigned during the first week of semester. First talks will be in the third week.
Goal
The goal of the seminar is not only to familiarize students with exciting new research topics, but also to teach basic scientific writing and oral presentation skills. The seminar will have a different structure from regular seminars to encourage more discussion and a deeper learning experience.
Seminars often suffer from poor attention retention and low student engagement. This is often due to the format of the seminar where only one student reads papers in-depth and then prepares a long presentation about one or sometimes several papers. There is little reason for the other students to really pay attention or engage in the discussion. To improve this the seminar will use a case-study format where all students read the same paper each week but fulfill different roles and hence prepare with different viewpoints in mind.
Student roles:- Historian: Find out how this paper sits in the context of the related work. Use bibliography tools to find the most influential papers cited by this work and at least one paper influenced by the work (and summarize the two papers).
- Presenter: Give a short talk about the paper that you read in depth.
- Reviewer: (e.g., reviewer of UIST/ICML/PLDI ): Complete a full critical review of the paper. Use the original review from and come to a recommendation whether the paper should be accepted or not.
- PhD student: Propose a follow-up project for your own research based on this paper - importantly the project should be directly inspired by the paper or even use/extend the method proposed.
- All students (every week): Come up with alternative title; find a missing result that the paper should have included.
Guidelines on how to read and review a paper can be found here. Please be aware that the reviewing guidelines are more tailored to Interactive Systems papers.
Schedule
| Wk. | Date | TA | Historian | Presenter | Reviewer | PhD Student | Paper |
|---|---|---|---|---|---|---|---|
| 1 | 05.10.2016 | Stefan Stevšić | Rudi David | Permenev Anton | Chang Li | Maximova Alexandra | A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning |
| 1 | 05.10.2016 | Petar Tsankov | Bucher Dominik Christoph | Maximova Alexandra | Keyes Daniel | Verhulst Tobias Florian | Estimating Types in Binaries using Predictive Modeling |
| 2 | 12.10.2016 | Fabrizio Pece | Keyes Daniel | Agrawal Rishu | Ghosh Partha | Permenev Anton | Online Detection and Classification of Dynamic Hand Gestures with Recurrent 3D Convolutional Neural Networks |
| 2 | 12.10.2016 | Pavol Bielik | Koller Zeno | Rumen Paletov | Verhulst Tobias Florian | Bucher Dominik Christoph | Explaining and harnessing adversarial examples |
| 3 | 19.10.2016 | Dimitar Dimitrov | Verhulst Tobias Florian | Prangishvili Iveri | Permenev Anton | Buffat Rene Michael | Static Analysis for Probabilistic Programs: Inferring Whole Program Properties from Finitely Many Paths |
| 3 | 19.10.2016 | Stefan Stevšić | Rumen Paletov | Goebel Fabian | Maximova Alexandra | Prangishvili Iveri | Deep spatial autoencoders for visuomotor learning |
| 4 | 26.10.2016 | Petar Tsankov | Prangishvili Iveri | Koller Zeno | Agrawal Rishu | Goebel Fabian | The Limitations of Deep Learning in Adversarial Settings |
| 4 | 26.10.2016 | Jie Song | Goebel Fabian | Ghosh Partha | Rumen Paletov | Koller Zeno | Structural-RNN: Deep Learning on Spatio-Temporal Graphs |
| 5 | 02.11.2016 | Jie Song | Ghosh Partha | Verhulst Tobias Florian | Goebel Fabian | Rumen Paletov | Conditional Random Fields as Recurrent Neural Networks |
| 6 | 09.11.2016 | Emre Aksan | Buffat Rene Michael | Keyes Daniel | Rudi David | Koller Zeno | Procedural Modeling Using Autoencoder Networks |
| 7 | 16.11.2016 | Pavol Bielik | Permenev Anton | Rudi David | Buffat Rene Michael | Agrawal Rishu | Learning Simple Algorithm From Examples |
| 7 | 23.11.2016 | Emre Aksan | Maximova Alexandra | Bucher Dominik Christoph | Prangishvili Iveri | Keyes Daniel | Weakly-supervised Disentangling with Recurrent Transformations for 3D View Synthesis |
| 8 | 23.11.2016 | Timon Gehr | Agrawal Rishu | Buffat Rene Michael | Bucher Dominik Christoph | Ghosh Partha | Unsupervised Learning by Program Synthesis |