DeepTrip

Abstract

In this work we propose DeepTrip -an end-to-end method for better understanding of the underlying human mobility and improved modeling of the POIs' transitional distribution in human moving patterns. DeepTrip consists of: a Trip Encoder to embed a given route into a latent variable with a recurrent neural network (RNN); and a Trip Decoder to reconstruct this route conditioned on an optimized latent space. Simultaneously, we define an Adversarial Net composed of a generator and critic, which generates a representation for a given query and uses a critic to distinguish the trip representation generated from Trip Encoder and query representation obtained from Adversarial Net. DeepTrip enables regularizing the latent space and generalizing users' complex check-in preference. We demonstrate the effectiveness and efficiency of the proposed model, and the experimental evaluations show that DeepTrip outperforms the state-of-the-art baselines on various evaluation metrics.

Deep-Reinforcement-Learning-Based Planner for City Tours for Cruise Passengers

“…This is why it is crucial to consider that the selection of each PoI heavily impacts the selection of the successive PoIs since the time spent visiting the computed sequence of PoIs drives the choice of the rest of the tour to meet the time constraints imposed by the cruise departure. Approaches that rely on immediate rewards for selecting each PoI [20] are not feasible in our case since long-term rewards should be considered. In order to take into account long-term rewards, a reinforcement learning approach that considers these variable constraints is proposed with the purpose of limiting the negative effects of tourist itineraries of cruise passengers in the destination city.…”