Title: Benefits and Perspectives of Automated Algorithm Selection for the Traveling Salesperson Problem
Begin: Thursday, 22. of October 2020 (4:15 PM)
via Zoom Meeting
Abstract:
The Euclidean Traveling Salesperson Problem (TSP) is a classical optimization problem which is of high relevance for science and industry. Although it has been well-studied for decades, there is no algorithm in the class of inexact TSP optimization that is superior to all its competitors. Consequently, choosing the "right" algorithm for optimizing a given TSP instance is a challenging task in itself, and choosing the "wrong" algorithm can have severe implications on the overall performance. In recent years, automated algorithm selection has proven to be a very effective method to address this challenge in an automated way, thus improving the state of the art in this particular optimization domain. In my presentation, I will summarize the status quo of automated algorithm selection for the TSP and present some research perspectives in times of automation, digitization and big data.
