Reasoning and LearningThe group is interested broadly in the following topics:
Stream Reasoning and Learning - Research on how to combine logic-based stream reasoning, incremental reasoning over streaming inputs, and learning-based probabilistic predictive models in order to reason over both incomplete and uncertain observations as well as uncertain predictions of the future.
Combinatorial Assignment - The study of how to distribute bundles of indivisible elements (e.g., goods, agents, sensors) among a set of alternatives (e.g., buyers, tasks, targets) to maximize the aggregated expected utility. This is a central problem in artificial intelligence, operations research, and algorithmic game theory; with applications in for example task/resource allocation, winner determination for combinatorial auctions, multi-target tracking and team/coalition formation.
Reinforcement Learning - The study of learning of how to act through interactions with an environment. We are especially interested in multi-agent and multi-objective reinforcement learning where multiple agents learn to act to maximize several independent objectives. We also study verification and validation of policies to allow reinforcement learning to be used in safety critical applications.
Time-series Learning - The study of how to learn models, often generative, based on time series data. We are especially interested in learning generative trajectory models for applications in autonomous systems, logistics and mobility.
Synthetic Data Generation - The study of how to generate tailored synthetic data either from data that cannot be used directly due to privacy concerns, bias or other issues or generative models such as GANs or simulators.
Reasoning and Learning - The study of principled approaches to integrate or combine reasoning and learning.
The research is funded by Knut and Alice Wallenberg Foundation (KAW), Wallenberg AI, Autonomous Systems and Software Program (WASP), Marcus and Amalia Wallenberg Foundation (MAW), WASP Humanities and Society (WASP-HS), Vinnova, Horizon 2020, ELLIIT, Trafikverket, Swedish National Graduate School in Computer Science (CUGS), and Center for Industrial Information Technology (CENIIT).
Page responsible: Patrick Doherty
Last updated: 2021-01-17