Case-based myopic reinforcement learning for satisfying target service level in supply chain

Kwon, Ick-Hyun; Kim, Chang Ouk; Jin, Jun; Lee, Jung Hoon

doi:10.1016/j.eswa.2007.07.002

Cited by 24 publications

(12 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, they don't show significantly better performance than linear regression. A case-based myopic reinforcement learning algorithm appropriate for the nonstationary inventory control problem of supply chain that has a large state space is proposed in [10]. The simulationbased experiment is performed to show the good performance of such algorithm.…”

Section: Learning Applications In Simulations Of Supply Chainsmentioning

confidence: 99%

Learning Framework for Multi-Agent Simulation of Supply Chain Systems

Jiang

Sheng

2008

2008 International Conference on Computer Science and Information Technology

View full text Add to dashboard Cite

Multi-agent System (MAS) is becoming a popularand useful method applied to solve problems in supply chains under dynamic environment. The learning ability is one of the most important components for "intelligent" agents. It helps agents adjust to dynamic environment more quickly and gain experience for making decisions. This paper proposes a learning framework for multi-agent simulation of supply chain systems. And this framework is explained step by step in 4 levels: individual, department, organization and supply chain system. The validity of such framework is established by simulating an example learning method using the simplified version of the framework. Furthermore, the application of such framework is not limited for SC, but also for community, government etc.

show abstract

Section: Learning Applications In Simulations Of Supply Chainsmentioning

confidence: 99%

Learning Framework for Multi-Agent Simulation of Supply Chain Systems

Jiang

Sheng

2008

2008 International Conference on Computer Science and Information Technology

View full text Add to dashboard Cite

show abstract

“…ADP has been recently introduced into inventory management research by Van Roy et al (1997), Godfrey et al (2001), Pontrandolfo, Gosavi, and Okobaa (2002), Giannoccaro and Pontrandolfo (2002), Shervais et al (2003), Kim et al (2005), Choi et al (2006), Topaloglu and Kunnumkal (2006), Iida and Zipkin (2006), Chaharsooghi et al (2008), Kim et al (2008), Kwon et al (2008) and Jiang and Sheng (2009). Fig.…”

Section: Literature Reviewmentioning

confidence: 99%

“…Jiang and Sheng (2009), Kim et al (2008), Kwon et al (2008), Kim et al (2005) used a look-up table to implement a cost-to-go approximation. A look-up table is a simple index table where each entry, an approximate cost, can be accessed by an index, a state-action pair.…”

Section: Function Approximations Used In Learning-based Adp Studiesmentioning

confidence: 99%

“…Their approach is similar to TD0, but it only approximates a current state-action value without a cost-to-go. Kwon et al (2008) and Jiang and Sheng (2009) used the casebased myopic reinforcement learning (CMRL) method developed by Kwon et al (2008) CMRL is based on a combination of an action-value method and a case-based reasoning technique. Case-based reasoning is state aggregation with an ability to create a new aggregation when an observed state is over a preset range of any existing aggregation group.…”

Section: Learning-based Adpmentioning

confidence: 99%

“…A simulation-based ADP method uses simulation, e.g., Monte Carlo simulation, to approximate state-action costs. Most previous studies of an ADP application to inventory management focused on learning-based ADP; see, e.g., Van Roy, Bertsekas, Lee, and Tsitsiklis (1997), Giannoccaro and Pontrandolfo (2002), Shervais, Shannon, and Lendaris (2003), Kim, Jun, Baek, Smith, and Kim (2005), Kwon, Kim, Jun, and Lee (2008), Kim, Kwon, and Baek (2008), Chaharsooghi, Heydari, and Zegordi (2008) and Jiang and Sheng (2009). Only Choi, Realff, and Lee (2006) investigated simulationbased ADP.…”

Section: Introductionmentioning

confidence: 98%

See 2 more Smart Citations

Approximate dynamic programming for an inventory problem: Empirical comparison

Katanyukul

Duff

Chong

2011

Computers & Industrial Engineering

View full text Add to dashboard Cite

Intelligent Inventory Control: Is Bootstrapping Worth Implementing?

Katanyukul

Chong

Duff

2012

Intelligent Information Processing VI

View full text Add to dashboard Cite

The common belief is that using Reinforcement Learning methods (RL) with bootstrapping gives better results than without. However, inclusion of bootstrapping increases the complexity of the RL implementation and requires significant effort. This study investigates whether inclusion of bootstrapping is worth the effort when applying RL to inventory problems. Specifically, we investigate bootstrapping of the temporal difference learning method by using eligibility trace. In addition, we develop a new bootstrapping extension to the Residual Gradient method to supplement our investigation. The results show questionable benefit of bootstrapping when applied to inventory problems. Significance tests could not confirm that bootstrapping had statistically significantly reduced costs of inventory controlled by a RL agent. Our empirical results are based on a variety of problem settings, including demand correlations, demand variances, and cost structures.

show abstract

Case-based myopic reinforcement learning for satisfying target service level in supply chain

Cited by 24 publications

References 14 publications

Learning Framework for Multi-Agent Simulation of Supply Chain Systems

Learning Framework for Multi-Agent Simulation of Supply Chain Systems

Approximate dynamic programming for an inventory problem: Empirical comparison

Intelligent Inventory Control: Is Bootstrapping Worth Implementing?

Contact Info

Product

Resources

About