Researchers present at leading artificial intelligence conference

School of Electrical Engineering and Computer Science (EECS) researchers recently participated in the Association for the Advancement of Artificial Intelligence (AAAI) Conference on Artificial Intelligence. The premier conference, held this year in Vancouver BC, promotes research in artificial intelligence and brings together researchers, scientists, students, and engineers in AI disciplines for scientific exchange.

Faculty and students in EECS presented several papers, covering a variety of topics including AI methods for safe decision-making, optimization of expensive functions for science and engineering applications, and deep models for snow-water equivalent prediction. Their work included ways to combine models from related tasks, new adaptive experimental design algorithms, and how to intelligently search for high-performing inputs outside the training data.

Man standing by a research poster.
Aryan Deshwal (Ph.D. Candidate) stands next to his exhibit at the Association for the Advancement of Artificial Intelligence Conference on Artificial Intelligence.

EECS faculty members Jana Doppa and Kaiyan Qiu along with graduate student Aryan Deshwal organized a workshop on AI to Accelerate Science and Engineering with the theme of AI for materials and manufacturing.

Faculty members Ganapati Bhat, Yan Yan, Nghia Hoang, and Doppa along with graduate student Taha Belkhouja gave a tutorial on Advances in Robust Time-Series Machine Learning: From Theory to Practice.

Hoang was also selected for the New Faculty Highlights program and gave an invited talk while Doppa chaired the student abstract program.

Papers presented included:

  • Few-Shot Learning via Repurposing Ensemble of Black-Box Models (Nghia Hoang)
  • Safe Reinforcement Learning with Instantaneous Constraints: The Role of Aggressive Exploration (Honghao Wei)
  • Offline Model-based Black-Box Optimization via Policy-guided Gradient Search (Yassine Chemingui, Aryan Deshwal, Nghia Hoang, and Jana Doppa)
  • Pareto front Diverse Batch Multi-Objective Bayesian Optimization (Alaleh Ahmadianshalchi, Syrine Belakaria, and Jana Doppa)
  • Attention-based models for snow-water equivalent prediction (Krishu Thapa, Ananth Kalyanaraman)