Publication
Title
A new perspective on classification : optimally allocating limited resources to uncertain tasks
Author
Abstract
A central problem in business concerns the optimal allocation of limited resources to a set of available tasks, where the payoff of these tasks is inherently uncertain. Typically, such problems are solved using a classification framework, where task outcomes are predicted given a set of characteristics. Then, resources are allocated to the tasks predicted to be the most likely to succeed. We argue, however, that using classification to address task uncertainty is inherently suboptimal as it does not take into account the available capacity. We present a novel solution that directly optimizes the assignment's expected profit given limited, stochastic capacity. This is achieved by optimizing a specific instance of the net discounted cumulative gain, a commonly used class of metrics in learning to rank. We demonstrate that our new method achieves higher expected profit and expected precision compared to a classification approach for a wide variety of application areas.
Language
English
Source (journal)
Decision support systems. - Amsterdam
Publication
Amsterdam : 2024
ISSN
0167-9236
DOI
10.1016/J.DSS.2023.114151
Volume/pages
179 (2024) , p. 1-9
Article Reference
114151
ISI
001166382700001
Full text (Publisher's DOI)
Full text (open access)
The author-created version that incorporates referee comments and is the accepted for publication version Available from 19.07.2024
Full text (publisher's version - intranet only)
UAntwerpen
Faculty/Department
Research group
Project info
POSITE: Process optimization with sequential individual treatment effects.
Machine learning for fraud analytics.
Publication type
Subject
Affiliation
Publications with a UAntwerp address
External links
Web of Science
Record
Identifier
Creation 29.03.2024
Last edited 17.04.2024
To cite this reference