Скачать с ютуб Recent Advances in Deep Learning for Routing Problems | Chaitanya K. Joshi в хорошем качестве

Recent Advances in Deep Learning for Routing Problems | Chaitanya K. Joshi

Join the Learning on Graphs and Geometry Reading Group: https://hannes-stark.com/logag-readin... Blog post “Recent Advances in Deep Learning for Routing Problems": https://www.chaitjo.com/post/deep-lea... Abstract: Developing Graph Neural Network-driven solvers for combinatorial optimization problems such as the Travelling Salesperson Problem have seen a surge of academic interest recently. This talk aims to serve as a whirlwind tour of this research area. We first presents a Neural Combinatorial Optimization pipeline that unifies several recently proposed models into one single framework. Through the lens of the pipeline, we analyze recent advances in deep learning for routing problems and provide new directions to stimulate future research. Authors: Chaitanya K. Joshi, Rishabh Anand Twitter Hannes:   / hannesstaerk   Twitter Dominique:   / dom_beaini   Twitter Valence Discovery:   / valence_ai   Reading Group Slack: https://join.slack.com/t/logag/shared... Chapters ~ 00:00 - Intro 01:16 - Overview/agenda 02:03 - Combinatorial Optimization 06:36 - Deep Learning for Routing Problems 08:19 - Neural Combinatorial Optimization 13:22 - Unified Pipeline for Neural Combinatorial Optimization 40:28 - Characterizing Prominent Papers via the Pipeline 45:20 - Case Studies of Recent Advances & Future Work 47:25 - Leveraging Equivariance and Symmetries 57:29 - Improved Graph Search Algorithms 1:07:04 - Learning to Improve Sub-Optimal Solutions 1:12:28 - Improved Evaluation Protocols 1:14:39 - Understudied Combinatorial Problems 1:20:44 - Q+A