Mixed Logit Model Better Predicts Travel Mode Preference with Varied Distributions.
The researchers looked at different ways to use a model to predict how people choose different modes of transportation. They compared two models and three different ways to guess the random parts of the model. They found that one model was better at fitting the data and that the three different ways to guess the random parts worked better for different things. They also found that the better model was better at showing what people like when they choose how to travel.
