https://zoom.us/j/92295392734?pwd=K2lwNU1LWHdxRGpia0pMcExMcmVwdz09
Meeting ID: 922 9539 2734 Passcode: 790476
Paper: Demystifying Black-box Models with Symbolic Metamodels by A. Alaa, M. V. D. Schaar Proposer: Daniel Kienitz
Paper: KoGuN: Accelerating Deep Reinforcement Learning via Integrating Human Suboptimal Knowledge Proposer: Fazl Barez
Paper: Informed Machine Learning – Towards a Taxonomy of Explicit Integration of Knowledge into Machine Learning Task: How can we extend this to first principles (i.e. Show mathematically how to incorporate rules/constarints in ML pipline?) Proposer: Fazl Barez
- The reading group members propose a research paper to read and discuss. It can be a paper that interests someone for their own research, but it is advisable that in addition that paper is of broader interest to the group. (lets establish the latter in each case individually using this slack channel)
- The reading group members commit to attend regularly -- to be fair and collegiate to other proposers.
- Each member of the reading group makes a reasonable effort to read parts specified by the proposer before the meeting. (and the proposer must be mindful of the proposed load)
- The reading group slot is 1 hour, during which the proposer is expected to guide the discussion of the proposed paper, and the rest of the group contribute their thoughts/opinions. Slides may be used as a guide, but mostly it is expected that the group reads and discusses the proposed text.
-
The reading group is a social exercise in good paper writing (not just reading!), as we all learn through best practices and discussions. Hence, rule No 5 is to have fun and enjoy it.