Introduction and Research Fellows Lightning Talks
summary
Five-minute snapshots present concrete claims and tools: queueing work gives implementable rules (how many hardware priority buckets and handling unknown sizes) and theory via SOAP and WINE that yields practical SRPT variants; fairness work quantifies the short-term cost of exploration versus long-term diversity; participatory budgeting proves constant-factor approximate core existence; graph methods prove non-backtracking walks beat simple walks for clustering; physics ML fixes antisymmetrized wavefunctions; TransNet model links epidemic/neuronal dynamics to learnable networks; a constant-sample epidemic estimator predicts outbreaks without full networks.
AI-written summary · about
chapters (13)
- 0:00Welcome to Simons Industry Day
- 1:25Schedule overview
- 3:08Queuing and scheduling theory
- 8:45Algorithmic fairness through exploration
- 13:51Fair provision of public goods
- 17:11Incentive-aware machine learning
- 22:32Symmetries in neural networks
- 27:55Clustering in large networks
- 32:19Algorithmic game theory
- 36:32Transmission neural networks
- 41:19Graph neural network geometry
- 46:15Epidemic prediction
- 48:50Closing remarks
related talks
56mLightning Talks!
47mLightning talks - session 1
45mLIGHTNING TALKS
01h 07mIndustry Presentations
01h 12mA Brief Introduction to Algorithms, Game Theory and Risk-Averse Decision Making
54mLightning Talks + Closing Remarks DAY 1
36mPyData London 2017, Lightning Talks - Saturday
01h 16mLightning Talks (01)