A Generalization Performance Improvement Approach for Reinforcement Learning-Based Multidimensional Knapsack Problem Solving 


Vol. 48,  No. 11, pp. 1418-1428, Nov.  2023
10.7840/kics.2023.48.11.1418


PDF
  Abstract

The Knapsack Problem, one of the combinatorial optimization problems, has no known method for finding the optimal solution within polynomial time. Potential applications of solutions to this problem include logistics and warehouse management, manufacturing and production planning, resource allocation, and scheduling. Recently, attempts have been made to solve the Knapsack Problem using reinforcement learning. However, many proposed approaches are dependent on the number of items in the problem, requiring individual model training for each set of items. This paper proposes a Markov Decision Process and a neural network structure that exploits the scale-invariance of the Knapsack Problem and is independent of the number of items. As a result, the method proposes and tests the performance of an approach applicable to the extended Multi-dimensional Knapsack Problem, independent of the number of items. The method's generalization performance, a key strength, is evaluated to demonstrate its scalability, reusability, and generality.

  Statistics
Cumulative Counts from November, 2022
Multiple requests among the same browser session are counted as one view. If you mouse over a chart, the values of data points will be shown.


  Related Articles
  Cite this article

[IEEE Style]

Y. Choi, Y. Seok, J. Kim, Y. Han, "A Generalization Performance Improvement Approach for Reinforcement Learning-Based Multidimensional Knapsack Problem Solving," The Journal of Korean Institute of Communications and Information Sciences, vol. 48, no. 11, pp. 1418-1428, 2023. DOI: 10.7840/kics.2023.48.11.1418.

[ACM Style]

Yohan Choi, Yeong-Jun Seok, Ju-Bong Kim, and Youn-Hee Han. 2023. A Generalization Performance Improvement Approach for Reinforcement Learning-Based Multidimensional Knapsack Problem Solving. The Journal of Korean Institute of Communications and Information Sciences, 48, 11, (2023), 1418-1428. DOI: 10.7840/kics.2023.48.11.1418.

[KICS Style]

Yohan Choi, Yeong-Jun Seok, Ju-Bong Kim, Youn-Hee Han, "A Generalization Performance Improvement Approach for Reinforcement Learning-Based Multidimensional Knapsack Problem Solving," The Journal of Korean Institute of Communications and Information Sciences, vol. 48, no. 11, pp. 1418-1428, 11. 2023. (https://doi.org/10.7840/kics.2023.48.11.1418)
Vol. 48, No. 11 Index

Search
(IN TITLE, AUTHOR, ABSTRACT,KEYWORDS)

Advanced Search ▼

POPULAR KEYWORDS
(TOP 10 KEYWORDS)

Recent Publications
(LAST 3 YEARS)






pISSN : 1226 - 4717
eISSN : 2287 - 3880