Discrete Choice Model in the Big Data Era

- Creating Abundant Datasets about Choice Situations -

抄録

The discrete choice model (DCM) is a promising method used to predict the choices of decision-makers. The most basic DCM is a multinomial logit (MNL) model. One of the basic characteristics of this model is that it exhibits independence from irrelevant alternatives (IIA). This property can be seen as a restriction imposed by the model. However, we are now entering a new era where abundant data about choice situations are available. In this era of “big data,” we propose use of a simple MNL model to overcome the problem of IIA by considering interaction effects between the attributes of decision-makers and attributes of alternatives.