based on features of the potential hotel customer, forming a plurality of clusters using a cluster model, each cluster comprising a corresponding weight and cluster probabilities, each cluster corresponding to a different hotel customer persona;

generating an initial estimated mixture of multinomial logit (MNL) models, each comprising a choice model and corresponding to each of the plurality of clusters, each MNL model based on parameters related to hotel room offerings and configured to predict a choice probability of room categories and rate code combinations for the potential hotel customer, the mixture of MNL models comprising a weighted likelihood function based on the features and the weights;

forming an Expectation Maximation functionality comprising a weighted likelihood function comprising a combination of the cluster model and the choice model, the weighted likelihood function comprising an objective function comprising a plurality of model parameters;

estimating the model parameters by finding a maximizer of the Expectation Maximation functionality comprising a plurality of iterations, the estimating comprising dynamically updating the cluster probabilities at each iteration evaluated at a value of the model parameters at a previous iteration and estimating an updated mixture of MNL models based on the updated cluster probabilities at each iteration; and

based on the value of the model parameters, generating the demand model.