Responsible AI for Decision Making

Goals

Develop AI tools that enable responsible decision-making in engineering and healthcare, with a focus on user-inspired research projects.

Issues Involved or Addressed

The advent of AI has transformed numerous fields by enabling efficient analysis of vast and complex datasets. However, integrating AI into decision-making processes poses challenges, particularly in high-stakes areas where incorrect decisions can have serious consequences. This emphasizes the need for trustworthy AI tools that are confidence-aware, robust and reliable, interpretable and explainable, and adaptable to domain shifts. Our objective is to develop trustworthy AI methods that support decision-making across diverse domains. This includes: Specialized Language Models: Tailoring language models to extract domain-specific knowledge, improving the efficiency and accuracy of information retrieval. Multi-modal Data Analysis: Creating AI systems capable of integrating data from various sources (e.g., text, images, sensors) to provide comprehensive insights for complex decisions. Time Series Forecasting: Developing robust models to predict future trends using historical data, aiding in proactive decision-making. AI-based Optimization: Leveraging AI to solve complex optimization problems, enhancing operational efficiency. These tools will be applied across sectors such as supply chain, manufacturing, e-commerce, mobility, power systems, and healthcare. Our research, driven by industrial case studies, aims to create significant societal impact.

Partners/Sponsors

NA