Research Seminar: Artificial Intelligence and Machine Learning

Welcome!

The seminar covers various topics on Artificial Intelligence (AI) and Machine Learning (ML) and is closely related to the research interests of the members of the Artificial Intelligence and Machine Learning Group.  

Many sessions take place in the form of a reading group, where we discuss the selected paper(s) on topics related to AI and ML. Although for each session there is a dedicated speaker, everyone should read the paper(s) before the meeting to be able to actively participate in the discussions. Additional reading material will be provided, covering the necessary background for each session. 

Typically the speakers are members of the group, sometimes we have also guest speakers (and participants).

The desired outcome of the seminar is the active participation of the students in scientific discussions and  and of course, a better understanding of AI/ML concepts, methods and algorithms and how can they be applied to particular problems and learning challenges.

Schedule

The seminar will take place on Thursdays from 11:00-13:00 via Webex: permanent meeting-link: https://fu-berlin.webex.com/fu-berlin-en/j.php?MTID=mf923fbb469747ce31097ba2cbbb19d11;  Meeting number: 121 5104834;  Password: JuRBiExq342

Talks schedule

A tentative plan of the talks for this semester, is as follows: 

1. Thursday 15/4/2021
   • Speaker: Arjun Roy 
   • Paper presentation (NeurIPS 2020): How Do Fair Decisions Fare in Long-term Qualification?
2. Thursday 22/4/2021
   • Speaker: Siamak Ghodsi
   • Paper presentation (FAT 2021): Towards Fair Deep Anomaly Detection
3. Thursday 29/4/2021 
   • Speaker: Simone Fabbrizzi
   • Paper presentation (arXiv):  Rethinking Content and Style:
     Exploring Bias for Unsupervised Disentanglement
4. Thursday 6/5/2021 
   • Speaker: Vishnu Unnikrishnan
   • Paper presentation (DSAA 2018): Leveraging entity similarities to develop personalised predictors
5. Thursday 20/5/2021 
   • Speaker: Eleni Ilkou
   • Paper presentation (arXiv):  Measuring social bias in knowledge graph embeddings
6. Thursday 27/5/2021 
   • Speaker: Tai Le Quy
   • Paper presentation (LAK 2020):  Constructing and Predicting School Advice for Academic Achievement - A Comparison of Item Response Theory and Machine Learning Techniques
7. Thursday 3/6/2021 
   • Speaker:  Philip Naumann
   • Paper presentation (Journal of Heuristics):  Biased random-key genetic algorithms for combinatorial optimization
8. Thursday 10/6/2021 
   • Speaker:  Simone Fabbrizzi
   • Paper presentation (Eurographics Symposium on Point-Based Graphics 2007):  The Mapper algorithm (Topological Methods for the Analysis of High Dimensional Data Sets and 3D Object Recognition)
9. Thursday 17/6/2021 
   • Speaker:  Siamak Ghodsi
   • Topic presentation (IJCAI and KDD papers): FAHT-ADWIN
10. Thursday 24/6/2021 
    • Speaker: Arjun Roy & Siamak Ghodsi
    • Topic presentation : Multifairness and MOO for fairness
11. Thursday 1/7/2021 
    • Speaker:  Yi Cai
    • Paper presentation (ACL 2020): Generating Hierarchical Explanations on Text Classification via Feature Interaction Detection
12. Thursday 8/7/2021 
    • Speaker:  Philip Naumann
    • Topic presentation:  Counterfactual generation
13. Thursday 15/7/2021 
    • Speaker:  Tai Le Quy
    • Topic presentation (EDM 2011 ICML 2019): Spectral clustering in educational data mining
      Guarantees for Spectral Clustering with Fairness Constraints

Literature

A good understanding of basic AI/ML models and algorithms is expected.

For each topic, further reading material will be provided in advance (check Resources section) which can help you to prepare for the sessions (including preparing your questions for the speakers).

Still, if some concepts are unknown or not well understood, we can discuss them during the seminar.