Sasin Research Seminar - From Prediction to Decision: Why Model Value Comes from Meaningful Interpretation

REGISTER NOW

You are cordially invited to join Sasin Research Seminar.

From Prediction to Decision: Why Model Value Comes from Meaningful Interpretation

By: Napol Rachatasumrit, Ph.D. AI Researcher, Edsy Date: Friday, February 13, 2026 Time: 12.00 – 13.00 Venue: Room 201 at Sasin School of Management or online via Zoom Register here to reserve your seat Abstract: The conventional wisdom in data mining is that better models fit data better, but this overlooks a critical point: the value of machine learning models lies not only in prediction, but in how they are used to inform decisions. Models optimized for accuracy often fail to produce interpretations that are scientifically meaningful or practically actionable, an issue that becomes important in high-stakes, human-centered domains. I introduce the concept of meaningful models: inherently interpretable models, not post-hoc explanations, that generate decision-relevant insight. Through examples from educational analytics, I show how parameter confounding turns accurate models into ambiguous decision tools, and how redesign restores actionable insight. This advances a shift from predictive performance to decision usefulness, with implications for organizational decision systems beyond education.   For more information please contact +66-2218-4000 ext. 84095 or [email protected].