Human-Centered Machine Learning: A Statistical and Algorithmic Perspective
Building artificial intelligence systems from a human-centered perspective is increasingly urgent, as large-scale machine learning systems ranging from personalized recommender systems to language and image generative models are deployed to interact with people daily. In this thesis, we propose a guideline for building these systems from a human-centered perspective. Our guideline contains three steps: (i) identifying the role of the people of interest and their core characteristics concerned in the learning task; (ii) modeling these characteristics in a useful and reliable manner; and (iii) incorporating these models into the design of learning algorithms in a principled way.
We ground this guideline in two applications: personalized recommender systems and decision-support systems. For recommender systems, we follow the guideline by (i) focusing on users’ evolving preferences, (ii) modeling them as dynamical systems, and (iii) developing efficient online learning algorithms with provable guarantees to interact with users sharing different preference dynamics. For decision-support systems, we (i) choose decision-makers’ risk preferences to be the core characteristics of concern, (ii) model them in the objective function of the system, and (iii) provide a general procedure with statistical guarantees for learning models under diverse risk preferences. We conclude by discussing the future of human-centered machine learning and the role of interdisciplinary research in this field.
Open Philanthropy AI Fellowship
- Machine Learning
- Doctor of Philosophy (PhD)