Results 1  10
of
14
Combining Psychological Models with Machine Learning to Better Predict People’s Decisions
"... Creating 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 behavio ..."
Abstract

Cited by 8 (8 self)
 Add to MetaCart
Creating 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. One approach creates apriori predictions about people’s behavior, either based on theoretical rational behavior or based on psychological models, including bounded rationality. A second type of approach focuses on creating models based exclusively on observations of people’s behavior. At the forefront of these types of methods are various machine learning algorithms. This paper explores how these two approaches can be compared and combined in different 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, the exact action predicted by psychological models is not even clear, and machine learning models are even less accurate. Nonetheless, we present a novel approach of creating hybrid methods that incorporate features from psychological models in conjunction with machine learning in order to create significantly improved models for predicting people’s decisions. To demonstrate these claims, we
Discriminating among probability weighting functions using adaptive design optimization
 Journal of Risk and Uncertainty
, 2013
"... Probability weighting functions relate objective probabilities and their subjective weights, and play a central role in modeling choices under risk within cumulative prospect theory. While several different parametric forms have been proposed, their qualitative similarities make it challenging to di ..."
Abstract

Cited by 5 (4 self)
 Add to MetaCart
(Show Context)
Probability weighting functions relate objective probabilities and their subjective weights, and play a central role in modeling choices under risk within cumulative prospect theory. While several different parametric forms have been proposed, their qualitative similarities make it challenging to discriminate among them empirically. In this paper, we use both simulation and choice experiments to investigate the extent to which different parametric forms of the probability weighting function can be discriminated using adaptive design optimization, a computerbased methodology that identifies and exploits model differences for the purpose of model discrimination. The simulation experiments show that the correct (datagenerating) form can be conclusively discriminated from its competitors. The results of an empirical experiment reveal heterogeneity between participants in terms of the functional form, with two models (Prelec2, Linear in Log Odds) emerging as the most common bestfitting models. The findings shed light on assumptions underlying these models.
A Bayesian latent group analysis for detecting poor effort in the assessment of malingering
, 2012
"... Abstract Despite their theoretical appeal, Bayesian methods for the assessment of poor effort and malingering are still rarely used in neuropsychological research and clinical diagnosis. In this article, we outline a novel and easytouse Bayesian latent group analysis of malingering whose goal is ..."
Abstract

Cited by 2 (1 self)
 Add to MetaCart
(Show Context)
Abstract Despite their theoretical appeal, Bayesian methods for the assessment of poor effort and malingering are still rarely used in neuropsychological research and clinical diagnosis. In this article, we outline a novel and easytouse Bayesian latent group analysis of malingering whose goal is to identify participants displaying poor effort when tested. Our Bayesian approach also quantifies the confidence with which each participant is classified and estimates the base rates of malingering from the observed data. We implement our Bayesian approach and compare its utility in effort assessment to that of the classic belowchance criterion of symptom validity testing (SVT). In two experiments, we evaluate the accuracy of both a Bayesian latent group analysis and the belowchance criterion of SVT in recovering the membership of participants assigned to the malingering group. Experiment 1 uses a simulation research design, whereas Experiment 2 involves the differentiation of patients with a history of stroke from coached malingerers. In both experiments, sensitivity levels are high for the Bayesian method, but low for the belowchance criterion of SVT. Additionally, the Bayesian approach proves to be resistant to possible effects of coaching. We conclude that Bayesian latent group methods complement existing methods in making more informed choices about malingering.
How Does Prospect Theory Reflect Heuristics' Probability Sensitivity in Risky Choice?
"... Abstract Two prominent approaches to describing how people make decisions between risky options are algebraic models and heuristics. The two approaches are based on fundamentally different algorithms and are thus usually treated as antithetical, suggesting that they may be incommensurable. Using cu ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
Abstract Two prominent approaches to describing how people make decisions between risky options are algebraic models and heuristics. The two approaches are based on fundamentally different algorithms and are thus usually treated as antithetical, suggesting that they may be incommensurable. Using cumulative prospect theory (CPT;
THEORETICAL REVIEW Using Bayesian hierarchical parameter estimation to assess the generalizability of cognitive models of choice
, 2014
"... Abstract To be useful, cognitive models with fitted parameters should show generalizability across time and allow accurate predictions of future observations. It has been proposed that hierarchical procedures yield better estimates of model parameters than do nonhierarchical, independent approache ..."
Abstract
 Add to MetaCart
(Show Context)
Abstract To be useful, cognitive models with fitted parameters should show generalizability across time and allow accurate predictions of future observations. It has been proposed that hierarchical procedures yield better estimates of model parameters than do nonhierarchical, independent approaches, because the formers ’ estimates for individuals within a group can mutually inform each other. Here, we examine Bayesian hierarchical approaches to evaluating model generalizability in the context of two prominent models of risky choice— cumulative prospect theory (Tversky&Kahneman, 1992) and the transferofattentionexchange model (Birnbaum & Chavez, 1997). Using empirical data of risky choices collected for each individual at two time points, we compared the use of hierarchical versus independent, nonhierarchical Bayesian estimation techniques to assess two aspects of model generalizability: parameter stability (across time) and predictive accuracy. The relative performance of hierarchical versus independent estimation varied across the different measures of generalizability. The hierarchical approach improved parameter stability (in terms of a lower absolute discrepancy of parameter values across time) and predictive accuracy (in terms of deviance; i.e., likelihood). With respect to test–retest correlations and posterior predictive accuracy, however, the hierarchical approach did not outperform the independent approach. Further analyses suggested that this was due to strong correlations between some parameters within both models. Such intercorrelations make it difficult to identify and interpret single parameters and can induce high degrees of shrinkage in hierarchical models. Similar findings may also occur in the context of other cognitive models of choice.
Journal of Mathematical Psychology
"... This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal noncommercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or sel ..."
Abstract
 Add to MetaCart
(Show Context)
This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal noncommercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit:
By
, 2011
"... Bayesian modeling has been influential in cognitive science. However, many psychological models of behavior have difficult or intractable likelihood functions. This poses a major problem for Bayesian inference, which requires a likelihood to provide estimates of the posterior distribution. In this d ..."
Abstract
 Add to MetaCart
Bayesian modeling has been influential in cognitive science. However, many psychological models of behavior have difficult or intractable likelihood functions. This poses a major problem for Bayesian inference, which requires a likelihood to provide estimates of the posterior distribution. In this dissertation, I provide a detailed examination of a new approach designed to avoid evaluating the likelihood. I provide an overview of current algorithms, and introduce a new algorithm for the estimation of the posterior distribution of the parameters of hierarchical models. I then apply the technique to two popular cognitive models of episodic memory. Finally, I show how the method can be used to perform model selection by means of mixture modeling. ii Acknowledgments I would like to thank • Scott Brown for telling me how to run C from R, without this, I would
Correspondence concerning this article should be addressed to:
"... The stopsignal paradigm is frequently used to study response inhibition. In this paradigm, participants perform a twochoice response time task where the primary task is occasionally interrupted by a stopsignal that prompts participants to withhold their response. The primary goal is to estimate t ..."
Abstract
 Add to MetaCart
(Show Context)
The stopsignal paradigm is frequently used to study response inhibition. In this paradigm, participants perform a twochoice response time task where the primary task is occasionally interrupted by a stopsignal that prompts participants to withhold their response. The primary goal is to estimate the latency of the unobservable stop response (stop signal reaction time or SSRT). Recently, Matzke, Dolan, Logan, Brown, and Wagenmakers (in press) have developed a Bayesian parametric approach that allows for the estimation of the entire distribution of SSRTs. The Bayesian parametric approach assumes that SSRTs are exGaussian distributed and uses Markov chain Monte Carlo sampling to estimate the parameters of the SSRT distribution. Here we present an efficient and userfriendly software implementation of the Bayesian parametric approach —BEESTS — that can be applied to individual as well as hierarchical stopsignal data. BEESTS comes with an easytouse graphical user interface and provides users with summary statistics of the posterior distribution of the parameters as well various diagnostic tools to assess the quality of the parameter estimates. The software is open source and runs on Windows and OS X operating systems. In sum, BEESTS allows experimental and clinical psychologists to estimate entire distributions of SSRTs and hence facilitates the more rigorous analysis of stopsignal data.
Edited by:
, 2012
"... What’s in a name: a Bayesian hierarchical analysis of the ..."
(Show Context)