On March 16th, 2022 Prof. Sihem Amer-Yahia, CNRS – University Grenoble Alpes (France), will hold a seminar on “Should We Store Or Retrain ML Models For Data Exploration?” in DEIB Seminar Room “Nicola Schiavoni” at 11.00 am and in live streaming on the WebEx.

Exploratory Data Analysis (EDA) is an iterative and often tedious process. Several strategies have been proposed to ease the burden on users in EDA ranging from stepwise to full-guidance approaches. Stepwise approaches rely on computing utility functions that determine the best action to take at each step. Full-guidance approaches rely on learning end-to-end exploration policies. The challenge of EDA today resides in evaluation that itself is due to the virtually endless purpose of EDA: are users looking for a needle in a haystack, taking a tour of the data, or are they feeling lucky? This talk will illustrate those challenges and discuss the question of whether we should store learned policies or retrain them when needed.

 

Sihem Amer-Yahia is a Silver Medal CNRS Research Director and Deputy Director of the Lab of Informatics of Grenoble. She works on exploratory data analysis and fairness in job marketplaces. Before joining CNRS, she was Principal Scientist at QCRI, Senior Scientist at Yahoo! Research and Member of Technical Staff at at&t Labs. Sihem is PC chair for SIGMOD 2023 and vice president of the VLDB Endowment. She currently leads the Diversity&Inclusion initiative for the database community.


The event will be held online by Webex at link


Please, note: Green Pass is needed to attend the event in presence.