Time: Wednesday 8-Jul-2020 23:30 (This is a past event.)
Motivation / Abstract
The talk will bridge between cognitive models used in behavioural economics and ML. Using cognitive model priors in ML may lead to a new generation of human choice prediction models. Specifically, Best Estimate and Sampling Tools (BEAST) leverages subjective weighing functions in favour of tackling deficiencies of expected utility and prospect theory used in economics. The two works discussed use the BEAST for human gamble choice predictions from two different perspectives in ML models and outline the future path in the field.
This talk gives an overview on 2 works regarding Behavioural economics x ML fields: "Psychological Forest: Predicting Human Behavior (Ido et al.)" and "Cognitive Model Priors for Predicting Human Desicions (Bourgin et al.)". The two works discussed use the BEAST for human gamble choice predictions, from two different perspectives in ML models, and outline the future path in the field.