This paper was published through Galoá and has a deposited DOI. To cite this paper, use one of the standards below:
In case you are one of the co-authors and want to register this paper in your Lattes, use the following code: doi > 10.59254/sbpo-2024-193370
If you've NEVER registered a DOI in your Lattes, check our tutorial!This work formalizes a special case of the strongly NP-Hard problem Precedence Constrained Knapsack Problem (PCKP). In this special case, the precedence constraints are restricted to paths within the same class of items, and all items have weight equal to one. This special case is named as Path-Precedence Constrained Knapsack Problem with Unit Weights (PPCKP-UW) and it has real-world applications, including a demand in logistics for solving it in real-time. This work proves that PPCKP-UW is solvable in polynomial-time, compares the computational performance of solving it using seven different approaches, including integer linear formulations, algorithms based on dynamic programming and shortest path, and applying pre/post-processing tasks to the instances. Ten new instances were generated and shared with the source codes, allowing the reproduction of all experiments made. Experimental results showed that the proposed approaches run faster than the conventional formulation based on PCKP.
Thiago Augusto de Oliveira Silva
Boa tarde!
Inicialmente gostaria de parabenizar pela apresentação. Fiquei na dúvida sobre o recorte temporal da decisão. É horizonte rolante ou problema é definitivo de forma estática?
Já pensaram em trabalhar o problema com incertezas? Já pensaram em utilizar aproximação de programação dinâmica para definir políticas sub-ótimas ?
Talvez o uso de ferramentas de aprendizado por reforço, em especial o PPO, pode ajudar a resolver o problema em um formato dinâmico e estocástico a partir de uma política pré-"otimizada" o que garante uma decisão mais rápida.
With nearly 200,000 papers published, Galoá empowers scholars to share and discover cutting-edge research through our streamlined and accessible academic publishing platform.
Learn more about our products:
This proceedings is identified by a DOI , for use in citations or bibliographic references. Attention: this is not a DOI for the paper and as such cannot be used in Lattes to identify a particular work.
Check the link "How to cite" in the paper's page, to see how to properly cite the paper
Cristiano Martins Monteiro
Boa tarde!
Obrigado pelos comentários e pelas perguntas!
Respondendo:
As implementações e instâncias do artigo estão disponíveis abertamente caso alguém tenha interesse em seguir a pesquisa. O link para download e definição dos algoritmos propostos estão descritos no corpo do artigo.