Talks and Poster Presentations (with Proceedings-Entry):
L. Kletzander, N. Musliu, K. Smith-Miles:
"Instance Space Analysis for a Personnel Scheduling Problem";
Talk: IJCAI 2019 - DSO Workshop,
Macao, China;
2019-08-11; in: "Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence, {IJCAI} 2019",
(2019),
1
- 8.
English abstract:
In order to analyse strengths and weaknesses of different solution methods, we use Instance Space Analysis to perform an in-depth evaluation on the Rotating Workforce Scheduling Problem. At first
we present a set of features aiming to describe hardness of instances. We create a new, more diverse set of instances based on a first analysis revealing gaps and possible extensions. The results of three algorithms on the extended instance set show different strong and weak areas in the instance space as well as a transition from feasible to infeasible instances including a more challenging area of instances at
this transition.
Related Projects:
Project Head Nysret Musliu:
ARTIOS
Created from the Publication Database of the Vienna University of Technology.