Combining Psychological Models with Machine Learning to Better Predict People's Decisions

Abstract

The public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Creating agents that pro¯ciently interact with people is critical for many applications. Towards creating these agents, models are needed that e®ectively predict people's decisions in a variety of problems. To date, two approaches have been suggested to generally describe people's decision behavior. These models could either be based on theoretical rational behavior, or psychological models such as those based on bounded rationality. A second approach focuses on creating ABSTRACTCreating agents that pro¯ciently interact with people is critical for many applications. Towards creating these agents, models are needed that e®ectively predict people's decisions in a variety of problems. To date, two approaches have been suggested to generally describe people's decision behavior. These models could either be based on theoretical rational behavior, or psychological models such as those based on bounded rationality. A second approach focuses on creating models based exclusively on observations of people's behavior. At the forefront of these type of methods are various machine learning algorithms. This paper explores how these two approaches can be compared and combined in di®erent types of domains. In relatively simple domains, both psychological models and machine learning yield clear prediction models with nearly identical results. In more complex domains, psychological or machine learning alone cannot accurately predict people's decisions. However, improved models can be created by using machine learning techniques to re¯ne parameters within psychological models. In the most complex domains, the exact action predicted by psychological models is not even clear, and machine learning models are even less accurate. Nonetheless, by creating hybrid methods that incorporate features from psychological models in conjunction with machine learning we can create signi¯cantly improved models for predicting people's decisions. To demonstrate these claims, we present a survey of previous and new results, taken from representative domains ranging from a relatively simple optimization problem, a more complex path selection domain, and complex domains of negotiation and coordination without communication. il, inonzu@ariel.ac.il, azariaa1@mail.biu.ac.il, sarit@cs.biu.ac.il Combining Psychological Models with Machine Learning 2 Combining Psychological Models with AbstractCreating agents that proficiently interact with people is critical for many applications. Towards creating these agents, models are needed that effectively predict people's decisions in a variety of problems. To date, two approaches have been suggested to generally describe people's decision behavior. These models could either be based on theoretical rational behavior, or psych...

Scaling up psychology via Scientific Regret Minimization

“…On a broader note, we hope to further the development of a synergistic correspondence between psychology and data science approaches in scientific modeling (Rosenfeld, Zuckerman, Azaria, & Kraus, 2012;Dwyer, Falkai, & Koutsouleris, 2018;Bourgin, Peterson, Reichman, Griffiths, & Russell, 2019). Cognitive science famously grew out of the intersection of six different fields (Gardner, 1987), but some have suggested that this revolution did not create the SCIENTIFIC REGRET MINIMIZATION 26 emergence of a new discipline (Lakatos, 1986;Miller, 2003;Núñez et al, 2019).…”

Using a Stochastic Agent Model to Optimize Performance in Divergent Interest Tacit Coordination Games

“…It has already been suggested that achieving successful human-machine collaboration requires the modeling of human behavior for predicting people's decisions [7,8]. It has been shown, for example, that information about drivers (i.e., driving style) improves the prediction models regarding the use of an automated assistive system [9].…”

Advances in Machine Learning for the Behavioral Sciences

“…In the political sciences, decision trees have been used for modeling international conflicts and international negotiation (Druckman et al, 2006). Rosenfeld et al (2012) Table 1 psychology, used random forests to predict future suicide attempts of patients.…”