A Brief Introduction to the Use of Machine Learning Techniques in the Analysis of Agent-Based Models

This paper gives a succinct introduction to some basic concepts imported from the fields of Machine and Statistical Learning that can be useful in the analysis of complex agent-based models (ABM). The paper presents some guidelines in the design of experiments. It then focuses on considering an ABM simulation as a computational experiment relating parameters with a response variable of interest, i.e. a statistic obtained from the simulation. This perspective gives the opportunity of using a supervised learning algorithm to fit the response with the parameters. The fitted model can be used to better interpret and understand the relation between the parameters of the ABM and the results in the simulation.

[1]  Antonio Criminisi,et al.  Decision Forests: A Unified Framework for Classification, Regression, Density Estimation, Manifold Learning and Semi-Supervised Learning , 2012, Found. Trends Comput. Graph. Vis..

[2]  Richard J. Beckman,et al.  A Comparison of Three Methods for Selecting Values of Input Variables in the Analysis of Output From a Computer Code , 2000, Technometrics.

[3]  Thomas Lengauer,et al.  Permutation importance: a corrected feature importance measure , 2010, Bioinform..

[4]  Forrest Stonedahl,et al.  The Complexities of Agent-Based Modeling Output Analysis , 2015, J. Artif. Soc. Soc. Simul..

[5]  D. Ruppert.  The Elements of Statistical Learning: Data Mining, Inference, and Prediction , 2004 .

[6]  M. Macy,et al.  FROM FACTORS TO ACTORS: Computational Sociology and Agent-Based Modeling , 2002 .

[7]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[8]  José Manuel Galán,et al.  Techniques to Understand Computer Simulations: Markov Chain Analysis , 2009, J. Artif. Soc. Soc. Simul..

[9]  Myong-Hun Chang.  Agent-based Modeling and Computational Experiments in Industrial Organization: Growing Firms and Industries in silico , 2011 .

[10]  Robert Tibshirani,et al.  The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd Edition , 2001, Springer Series in Statistics.

[11]  Bruce Edmonds,et al.  Errors and Artefacts in Agent-Based Modelling , 2009, J. Artif. Soc. Soc. Simul..

[12]  José Manuel Galán,et al.  Effect of Resource Spatial Correlation and Hunter-Fisher-Gatherer Mobility on Social Cooperation in Tierra del Fuego , 2015, PloS one.

[13]  Achim Zeileis,et al.  Bias in random forest variable importance measures: Illustrations, sources and a solution , 2007, BMC Bioinformatics.

[14]  José Manuel Galán,et al.  Economía Artificial: Métodos de Inspiración Social en la Resolución de Problemas Complejos (Artificial Economics: Social Inspired Methods for Solving Complex Problems) , 2014 .

[15]  Richard Simon,et al.  Bias in error estimation when using cross-validation for model selection , 2006, BMC Bioinformatics.

[16]  Trevor Hastie,et al.  An Introduction to Statistical Learning , 2013, Springer Texts in Statistics.

[17]  Daniela M. Witten,et al.  An Introduction to Statistical Learning: with Applications in R , 2013 .

[18]  Achim Zeileis,et al.  BMC Bioinformatics BioMed Central Methodology article Conditional variable importance for random forests , 2008 .

[19]  Zhenzhou Lu,et al.  Variable importance analysis: A comprehensive review , 2015, Reliab. Eng. Syst. Saf..