Toward Trustable Model-centric Sharing for Collaborative Machine Learning

An IDS flagship project funded by AI Singapore (Research)


 

Publications

The following is a list of publications from the project team, most of which are in the top 10% of S&T journals and conferences in the field as tracked in a list approved by NRF and/or in the Clarivate Analytics Journal Citation Reports.

Project Publications

Model-Contrastive Federated Learning

Probabilistic Sequential Shrinking: A Best Arm Identification Algorithm for Stochastic Bandits with Corruptions

Collaborative Bayesian Optimization with Fair Regret

Learning to Learn with Gaussian Processes

Practical One-Shot Federated Learning for Cross-Silo Setting

Challenges and Opportunities of Building Fast GBDT Systems

Thompson Sampling Algorithms for Cascading Bandits

Validation Free and Replication Robust Volume-based Data Valuation

A Unifying Theory of Thompson Sampling for Continuous Risk-Averse Bandits

Gradient Driven Rewards to Guarantee Fairness in Collaborative Machine Learning

Incentivizing Collaboration in Machine Learning via Synthetic Data Rewards

Federated Learning on Non-IID Data Silos: An Experimental Study


Other Publications

A Survey on Federated Learning Systems: Vision, Hype and Reality for Data Privacy and Protection

Near-Optimal Task Selection with Mutual Information for Meta-Learning

Value-at-Risk Optimization with Gaussian Processes

Trusted-Maximizers Entropy Search for Efficient Bayesian Optimization