The AI So-Go-Chi Center held its regular research seminar on July 27, 2026, in a hybrid format at the Global Connect Hub, Tohoku University. The seminar featured two presentations highlighting recent advances in trustworthy AI and data-driven decision support.

Presentation 1

AI for Science SPReAD Project: ReLEM: Development of Trustworthy Generative EEG AI through Explainability and Knowledge Integration

Assistant Professor Watanee Jearanaiwongkul presented the latest progress of the AI for Science SPReAD project, introducing the ReLEM framework for developing trustworthy generative EEG AI. The presentation discussed how explainable AI techniques and knowledge integration can improve the safety, transparency, and reliability of generative EEG models for scientific and healthcare applications.

Presentation 2

What-if Analysis for New Product Marketing Using Conditional Deep Generative Models for Tabular Data

Professor Tsukasa Ishigaki introduced a data-driven framework for supporting new product marketing through what-if analysis based on conditional deep generative models for tabular data. The presentation demonstrated how generative AI can simulate alternative business scenarios and support decision-making by estimating the potential outcomes of different marketing strategies.

The seminar concluded with an engaging discussion among participants on trustworthy AI, generative models, explainability, and practical applications of AI in scientific research and business analytics.