avigal

Mutual information (MI) quantifies the dependence between features and target variables and is widely used in feature selection for downstream tasks. However, what works globally often fails locally and different feature subsets may best fit different subgroups. The paper “MISFEAT: Feature Selection for Subgroups with Mutual Information Estimation“, co-authored by Bar Genossar, Thinh On, Md Mouinul Islam, Ben Eliav, Senjuti Basu Roy and Avigdor Gal, introduces a framework

The paper “A Rank-Based Approach to Recommender System’s Top-K Queries with Uncertain Scores”, co-authored by Coral Scharf, Avigdor Gal, Haggai Roitman, and Carmel Domshlak will be presented in #SIGMOD 2025. Thrilled to share this groundbreaking work that revolutionizes how we think about ranking in recommender systems! 🚀 For years, recommender systems have struggled with uncertain scores, often

The paper “An unsupervised machine learning approach to the spatial analysis of urban systems through neighbourhoods’ dynamics” is a result of a collaborative effort with Alon Sagi, Avigdor Gal, Dani Broitman, and Danny Czamanski. It tells the story of the dynamics that affect the social-environmental sustainability of UK’s urban system using unsupervised learning. Read about

CopycHats is an LLM-based agent that performs schema matching just like humans, allowing us to test revolving doors research questions that we cannot with humans. Matan Solomon presented this work, co-authored with Bar Genossar and Avigdor Gal as part of the hilda2024 workshop in SIGMOD. Link: https://dl.acm.org/doi/10.1145/3665939.3665963

Improving entity resolution while playing battleship? Yes, we can. The paper, co-authored by Bar Genossar, Roee Shraga, and Avigdor Gal, was presented in sigmod2024, offering an active learning approach to entity resolution, using spatial uncertainty over a vectorspace.