TY - GEN
T1 - Toward inverse generative social science using multi-objective genetic programming
AU - Vu, Tuong Manh
AU - Brennan, Alan
AU - Probst, Charlotte
AU - Strong, Mark
AU - Epstein, Joshua M.
AU - Purshouse, Robin C.
N1 - Funding Information:
The research reported in this paper was funded by the National Institute on Alcohol Abuse and Alcoholism of the National Institutes of Health under Award Number R01AA024443. The authors have no conflicts of interest. TMV designed the primitives and implemented the model discovery process; CP developed the social norms model; TMV, MS, AB and RP designed the model discovery process. JME and RCP led the research and its conceptual underpinnings. TMV, CP, JME and RCP wrote the manuscript. The authors would also like to acknowledge support from the wider R01AA024443 'CASCADE' project team.
Publisher Copyright:
© 2019 Copyright held by the owner/author(s). ACM ISBN 978-1-4503-6111-8/19/07.
PY - 2019/7/13
Y1 - 2019/7/13
N2 - Generative mechanism-based models of social systems, such as those represented by agent-based simulations, require that intraagent equations (or rules) be specified. However there are often many different choices available for specifying these equations, which can still be interpreted as falling within a particular class of mechanisms. Whilst it is important for a generative model to reproduce historically observed dynamics, it is also important for the model to be theoretically enlightening. Genetic programs (our own included) often produce concatenations that are highly predictive but are complex and hard to interpret theoretically. Here, we develop a new method - based on multi-objective genetic programming - for automating the exploration of both objectives simultaneously. We demonstrate the method by evolving the equations for an existing agent-based simulation of alcohol use behaviors based on social norms theory, the initial model structure for which was developed by a team of human modelers. We discover a trade-off between empirical fit and theoretical interpretability that offers insight into the social norms processes that influence the change and stasis in alcohol use behaviors over time.
AB - Generative mechanism-based models of social systems, such as those represented by agent-based simulations, require that intraagent equations (or rules) be specified. However there are often many different choices available for specifying these equations, which can still be interpreted as falling within a particular class of mechanisms. Whilst it is important for a generative model to reproduce historically observed dynamics, it is also important for the model to be theoretically enlightening. Genetic programs (our own included) often produce concatenations that are highly predictive but are complex and hard to interpret theoretically. Here, we develop a new method - based on multi-objective genetic programming - for automating the exploration of both objectives simultaneously. We demonstrate the method by evolving the equations for an existing agent-based simulation of alcohol use behaviors based on social norms theory, the initial model structure for which was developed by a team of human modelers. We discover a trade-off between empirical fit and theoretical interpretability that offers insight into the social norms processes that influence the change and stasis in alcohol use behaviors over time.
KW - Generative social science
KW - Genetic programming
KW - Multi-objective optimization
UR - http://www.scopus.com/inward/record.url?scp=85072305916&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85072305916&partnerID=8YFLogxK
U2 - 10.1145/3321707.3321840
DO - 10.1145/3321707.3321840
M3 - Conference contribution
C2 - 33083795
VL - 2019
T3 - Genetic and Evolutionary Computation Conference : [proceedings]. Genetic and Evolutionary Computation Conference
SP - 1356
EP - 1363
BT - GECCO 2019 - Proceedings of the 2019 Genetic and Evolutionary Computation Conference
PB - Association for Computing Machinery, Inc
T2 - 2019 Genetic and Evolutionary Computation Conference, GECCO 2019
Y2 - 13 July 2019 through 17 July 2019
ER -