In this series of AI short talks Morteza Haghir Chehreghani, Associate Professor at Data Science and AI division, Department of Computer Science and Engineering at Chalmers, will handle the subject: Adaptive Information Acquisition and Sequential Decision Making in AI.
The first series of CHAIR Spotlight Research talks have the theme “New Chalmers researchers on the spot!”, meaning that researchers that came to Chalmers in the last three years are giving these talks.
In many AI and machine learning tasks, we need to deal with the lack of sufficient supervision (e.g., labeled data and beyond). We study such problems in the general form of ’undersupervised machine learning’. In particular, we consider the optimal value of information problem in the context of adaptive information acquisition, where the goal is to sequentially select a set of tests with a minimal cost, so that one can efficiently make the best decision based on the observed outcomes. Existing algorithms are either heuristics with no guarantees, or scale poorly (with exponential run time in terms of the number of available tests). Moreover, these methods assume a known distribution over the test outcomes, which is often not the case in practice. We propose a sampling-based online learning framework to address the above issues. First, assuming the distribution over hypotheses is known, we propose a dynamic hypothesis enumeration strategy, which allows efficient information gathering with strong theoretical guarantees. We show that with sufficient amount of samples, one can identify a near-optimal decision with high probability. Second, when the parameters of the hypotheses distribution are unknown, we propose an algorithm that learns the parameters progressively via posterior sampling in an online fashion. We further establish a rigorous bound on the expected regret. We demonstrate the effectiveness of our approach on a real-world interactive troubleshooting application, and show that one can efficiently make high-quality decisions with low cost.
Morteza Haghir Chehreghani is Associate Professor of AI and Machine Learning at Chalmers University of Technology, Department of Computer Science and Engineering, Data Science and AI division. He holds a Ph.D. in Computer Science (AI/Machine Learning group) from ETH Zurich (2014). After the Ph.D., he joined Xerox Research as a Staff Research Scientist I and then Staff Research Scientist II, where he led several long-term projects on foundations of AI and the applications. After about four years and in 2018, he joined Chalmers University of Technology. His research is on foundations of AI, machine learning and data science, as well as real-world applications in transport, energy and life science.
About CHAIR Spotlight on Research
The aim is to increase awareness of AI at Chalmers between Chalmers researchers and AI experts in industry. In the seminars, speakers present an overview of their current research and thoughts for new research, ideas, challenges – anything they believe to be of interest for other researchers. The seminar is taking place online and is scheduled to contain 30 minutes of presentation and 15 minutes of discussion. The seminars are open to all and are free of charge. CHAIR Spotlight Research talks are taking place on Fridays 13:00-13:45.