Mitigating Dimensionality in 2D Rectangle Packing Problem under Reinforcement Learning Schema

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Kołodziejczyk, Waldemar, Kaleta, Mariusz
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866929511891730432
author Kołodziejczyk, Waldemar
Kaleta, Mariusz
author_facet Kołodziejczyk, Waldemar
Kaleta, Mariusz
contents This paper explores the application of Reinforcement Learning (RL) to the two-dimensional rectangular packing problem. We propose a reduced representation of the state and action spaces that allow us for high granularity. Leveraging UNet architecture and Proximal Policy Optimization (PPO), we achieved a model that is comparable to the MaxRect heuristic. However, our approach has great potential to be generalized to nonrectangular packing problems and complex constraints.
format Preprint
id arxiv_https___arxiv_org_abs_2409_09677
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Mitigating Dimensionality in 2D Rectangle Packing Problem under Reinforcement Learning Schema
Kołodziejczyk, Waldemar
Kaleta, Mariusz
Machine Learning
Optimization and Control
This paper explores the application of Reinforcement Learning (RL) to the two-dimensional rectangular packing problem. We propose a reduced representation of the state and action spaces that allow us for high granularity. Leveraging UNet architecture and Proximal Policy Optimization (PPO), we achieved a model that is comparable to the MaxRect heuristic. However, our approach has great potential to be generalized to nonrectangular packing problems and complex constraints.
title Mitigating Dimensionality in 2D Rectangle Packing Problem under Reinforcement Learning Schema
topic Machine Learning
Optimization and Control
url https://arxiv.org/abs/2409.09677