Beyond the Top Choice: MIT Researchers Modernize Century-Old Models to Better Predict Human Preferences
How can we better predict what people want? MIT researchers have provided a major upgrade to ‘random utility models,’ a psychological and economic framework nearly a century old. By moving beyond simple ‘first-choice’ analysis and incorporating the ‘power of three’—considering the top three preferences of individuals—the team has unlocked a much higher level of predictive accuracy.
This breakthrough allows for a more nuanced understanding of decision-making processes. Whether it is predicting which commute route a person will take or which product a customer will click on, this new model accounts for the subtle trade-offs humans make. By refining these foundational tools of data science, the research offers powerful new applications for industries ranging from transportation planning to digital marketing.