Causality and Machine Learning
Outils
DoWhy: A library for causal inference
mai 2021
As computing systems are more frequently and more actively intervening in societally critical domains such as healthcare, education and governance, it is critical to correctly predict and understand the causal effects of these interventions. Without an A/B test, conventional machine…
EconML: A Python Package for ML-Based Heterogeneous Treatment Effects Estimation
juillet 2019
EconML is a Python package for estimating heterogeneous treatment effects from observational data via machine learning. This package was designed and built as part of the ALICE project at Microsoft Research with the goal to combine state-of-the-art machine learning techniques…
Project Azua: Data efficient decision making
septembre 2021
Many modern AI algorithms are known to be data-hungry, whereas human decision-making is much more efficient. The human can reason under uncertainty, actively acquire valuable information from the world to reduce uncertainty, and make personalized decisions given incomplete information. How…
RobustDG
juillet 2020
RobustDG is a library of ML models that generalize to unseen domains and provides evaluation metrics on accuracy and multiple robustness metrics.
DiCE: A library for generating Diverse Counterfactual Explanations
novembre 2021
DiCE is a Python library that can generate counterfactual explanations for any machine learning classifier. Counterfactual explanations present «what-if» perturbations of the input such that an ML classifier outputs a different class for those perturbations than the original predicted class.…
Debiasing Item-to-Item Recommendations With Small Annotated Datasets Release
octobre 2020
Implementation of «Debiasing Item-to-Item Recommendations With Small Annotated Datasets» (RecSys ’20)
Longitudinal Tweet ID dataset for a selection of Health, Social, and Business Experiences
avril 2017
This data set consists of the tweet IDs collected for the propensity-score analysis of longitudinal social media messages posted by people who mention specific health, social and business domains. This data set accompanies the paper, “Distilling the Outcomes of Personal…