TY - CONF
T1 - Improving a model of human planning via large-scale data and deep neural networks
AU - Kuperwajs, Ionatan
AU - Schütt, Heiko
AU - Ma, Wei Ji
N1 - Funding Information:
do so. Humans display such systematic biases in many tasks, and the literature on these biases and how to model them This work was supported by Graduate Research Fellowship may be informative to structure the biases players show in number DGE183930 from the National Science Foundation, 4-in-a-row (Griffiths, Chater, Kemp, Perfors, & Tenenbaum, grant number IIS-1344256 from the National Science Foun-2010). Technically, such biases could be incorporated into the dation, and grant number R01MH118925 from the National planning model fairly easily by simply increasing the proba-Institutes of Health.
Funding Information:
This work was supported by Graduate Research Fellowship number DGE183930 from the National Science Foundation, grant number IIS-1344256 from the National Science Foundation, and grant number R01MH118925 from the National Institutes of Health.
Publisher Copyright:
© 2022 The Author(s). This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY)
PY - 2022
Y1 - 2022
N2 - Models in cognitive science are often restricted for the sake of interpretability, and as a result may miss patterns in the data that are instead classified as noise. In contrast, deep neural networks can detect almost any pattern given sufficient data, but have only recently been applied to large-scale data sets and tasks for which there already exist process-level models to compare against. Here, we train deep neural networks to predict human play in 4-in-a-row, a combinatorial game of intermediate complexity, using a data set of 10, 874, 547 games. We compare these networks to a planning model based on a heuristic function and tree search, and make suggestions for model improvements based on this analysis. This work provides the foundation for estimating a noise ceiling on massive data sets as well as systematically investigating the processes underlying human sequential decision-making.
AB - Models in cognitive science are often restricted for the sake of interpretability, and as a result may miss patterns in the data that are instead classified as noise. In contrast, deep neural networks can detect almost any pattern given sufficient data, but have only recently been applied to large-scale data sets and tasks for which there already exist process-level models to compare against. Here, we train deep neural networks to predict human play in 4-in-a-row, a combinatorial game of intermediate complexity, using a data set of 10, 874, 547 games. We compare these networks to a planning model based on a heuristic function and tree search, and make suggestions for model improvements based on this analysis. This work provides the foundation for estimating a noise ceiling on massive data sets as well as systematically investigating the processes underlying human sequential decision-making.
KW - behavioral modeling
KW - machine learning
KW - neural networks
KW - planning
KW - sequential decision-making
UR - http://www.scopus.com/inward/record.url?scp=85146418857&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85146418857&partnerID=8YFLogxK
M3 - Paper
AN - SCOPUS:85146418857
SP - 1190
EP - 1196
T2 - 44th Annual Meeting of the Cognitive Science Society: Cognitive Diversity, CogSci 2022
Y2 - 27 July 2022 through 30 July 2022
ER -