BriefGPT.xyz
Mar, 2022
结合强化学习和最优传输的旅行商问题
Combining Reinforcement Learning and Optimal Transport for the Traveling Salesman Problem
HTML
PDF
Yong Liang Goh, Wee Sun Lee, Xavier Bresson, Thomas Laurent, Nicholas Lim
TL;DR
利用熵正则化最优传输技术作为深度强化学习网络中的一层，以实现更快速的学习并在端到端训练期间强制执行分配约束和规定，从而对于解决组合优化问题的效率进行优化。
Abstract
The
traveling salesman problem
is a fundamental
combinatorial optimization
problem with strong exact algorithms. However, as problems scale up, these exact algorithms fail to provide a solution in a reasonable ti
→