We are pleased to share that a research paper authored by Dr. Reda Alami, Researcher at our AI and Digital Science Research Center (AIDRC), titled “Bayesian Change-Point Detection for Bandit Feedback in Non-stationary Environments”, has been accepted for publication during the 14th Asian Conference on Machine Learning (ACML 2022).
ACML is a leading international forum in the AI domain, bringing together researchers, engineers, and practitioners of Machine Learning and other AI-related fields to explore and disseminate the discoveries and innovations made in this area. Sponsored by Google and Microsoft Research, the conference encourages significant and novel submissions to foster the sharing of new ideas and cultivate a dynamic AI ecosystem.
The paper explored the facilitation of sequential decisions under uncertainty and the challenges faced in a non-stationary environment. In this context, the AI must be trained to make optimal decisions regardless of changes in its surroundings. The paper proposed a solution that combines two optimal algorithms: one for decision-making under uncertainty and the other for the sequential detection of abrupt changes, known as Bayesian-CPD-TS (standing for Bayesian Change-Point Detection with Thompson Sampling).
Speaking on his paper’s forthcoming publication, Dr. Alami said: “ACML is a highly reputed forum for AI professionals and the acceptance of this paper is an honor for myself and for TII, clearly underlining the superior quality of our research and the impact it is already having around the world. I look forward to attending the upcoming edition of the forum that runs in a hybrid format in Hyderabad, India in December.”
Dr. Alami’s research efforts, along with that of the AIDRC’s research team, build upon the next generation of lifelong reinforcement learning algorithms that can easily adapt to a multitude of environmental changes. The acceptance also marks a recognition of the research caliber of the UAE’s scientific community in formulating cutting-edge algorithms for the advancement of ML in the country and the wider region.
