Empirically-Derived Behavioral Rules in Agent-Based Models Using Decision Trees Learned from Questionnaire Data

摘要: With the increasing trend in exploring use of agent-based models empirical contexts, this paper reflects on decision trees learned from questionnaire data as behavioral for agents. Decision are machine learning algorithms most commonly used mining literature, especially smaller datasets where other techniques such Bayesian Networks cannot be applied. In modelling have advantage over some that results more transparent, and can critiqued by domain experts without a background computing or artificial intelligence. However, sensitive to way which they constructed, particularly with respect preprocessing. We describe processes were derived context model everyday pro-environmental behavior at work, comparing various preprocessing methods their differences.