A Gaussian-Prioritized Approach for Deploying Additional Route on Existing Mass Transportation with Neural-Network-Based Passenger Flow Inference

摘要: Multi-criteria path planning is an important combinatorial optimization problem with broad real-world applications. Finding the Pareto-optimal set of paths ideal for all requiring features time-consuming and unclear to obtain subset optimal efficiently multiple origin states in space. Meanwhile, due rise deep learning, hybrid systems computational intelligence thrive recent years. When facing non-monotonic data or heuristics derived from pretrained neural networks, most existing methods oneto-all fail find solution. We employ Gaussian mixture model propose a target-prioritized searching algorithm called Multi-Source Bidirectional Gaussian-Prioritized Spanning Tree (BiasSpan) solving this multicriteria route given constraints including range, must-visit vertices, number recommended vertices. Experimental results on mass transportation system Tainan Chicago cities show that BiasSpan outperforms comparative 7% 24% runs reasonable time compared state-of-art route-planning algorithms.