Reward shaping to improve the performance of deep reinforcement learning in perishable inventory management

2021 ◽  
Author(s):  
Bram J. De Moor ◽  
Joren Gijsbrechts ◽  
Robert N. Boute
Keyword(s):  
Reinforcement Learning ◽  
Inventory Management ◽  
Perishable Inventory ◽  
Reward Shaping

State-of-the-art literature review on inventory models for perishable products

2018 ◽  
Vol 15 (3) ◽  
pp. 306-346 ◽  
Author(s):  
Vaibhav Chaudhary ◽  
Rakhee Kulshrestha ◽  
Srikanta Routroy
Keyword(s):  
Inventory Management ◽  
Systemic Review ◽  
Current Status ◽  
Future Research ◽  
Inventory Models ◽  
Research Gaps ◽  
Content Type ◽  
Perishable Items ◽  
Perishable Inventory ◽  
The Future

PurposeThe purpose of this paper is to review and analyze the perishable inventory models along various dimensions such as its evolution, scope, demand, shelf life, replenishment policy, modeling techniques and research gaps.Design/methodology/approachIn total, 418 relevant and scholarly articles of various researchers and practitioners during 1990-2016 were reviewed. They were critically analyzed along author profile, nature of perishability, research contributions of different countries, publication along time, research methodologies adopted, etc. to draw fruitful conclusions. The future research for perishable inventory modeling was also discussed and suggested.FindingsThere are plethora of perishable inventory studies with divergent objectives and scope. Besides demand and perishable rate in perishable inventory models, other factors such as price discount, allow shortage or not, inflation, time value of money and so on were found to be combined to make it more realistic. The modeling of inventory systems with two or more perishable items is limited. The multi-echelon inventory with centralized decision and information sharing is acquiring lot of importance because of supply chain integration in the competitive market.Research limitations/implicationsOnly peer-reviewed journals and conference papers were analyzed, whereas the manuals, reports, white papers and blood-related articles were excluded. Clustering of literature revealed that future studies should focus on stochastic modeling.Practical implicationsStress had been laid to identify future research gaps that will help in developing realistic models. The present work will form a guideline to choose the appropriate methodology(s) and mathematical technique(s) in different situations with perishable inventory.Originality/valueThe current review analyzed 419 research papers available in the literature on perishable inventory modeling to summarize its current status and identify its potential future directions. Also the future research gaps were uncovered. This systemic review is strongly felt to fill the gap in the perishable inventory literature and help in formulating effective strategies to design of an effective and efficient inventory management system for perishable items.

Real-Time Perishable Inventory Management with Mid-Cycle Returns and Reordering

2017 ◽  
Author(s):  
Kebing Chen ◽  
Jing-Sheng Jeannette Song ◽  
Jennifer Shang ◽  
Tiaojun Xiao
Keyword(s):  
Real Time ◽  
Inventory Management ◽  
Perishable Inventory

Effect of Different Dimensions of Inventory Management of Fruits and Vegetables on Profitability of Retail Stores: An Empirical Study

2017 ◽  
Vol 19 (1) ◽  
pp. 99-110
Author(s):  
Lamay Bin Sabir ◽  
Jamal A. Farooquie
Keyword(s):  
Inventory Management ◽  
Fruits And Vegetables ◽  
Statistical Tests ◽  
Retail Store ◽  
Primary Objective ◽  
Chi Square ◽  
Perishable Inventory ◽  
Competitive Edge ◽  
Significant Relationships ◽  
Different Dimensions

In today’s challenging and competitive scenario, Indian retailers (organized sector) of fruits and vegetables need more dynamic strategies in order to provide customer satisfaction and retention. Purchasing, overstocking, stock-out, throw away, markdowns, etc. are different activities that are undertaken in a retail store selling perishable inventory, especially fruits and vegetables. These factors affect the profitability of the retail store, directly or indirectly; hence, proper control over these factors must be the primary objective of the retailer selling fruits and vegetables. This article aims to find out significant relationships within these parameters of inventory management so that retailers find it helpful in devising strategies for a better competitive edge. First, factors are identified, and then, statistical tests (chi-square and analysis of variance) are used to derive conclusions.

Plan-based reward shaping for multi-agent reinforcement learning

2016 ◽  
Vol 31 (1) ◽  
pp. 44-58 ◽  
Author(s):  
Sam Devlin ◽  
Daniel Kudenko
Keyword(s):  
Reinforcement Learning ◽  
Potential Function ◽  
Single Agent ◽  
Individual Agent ◽  
Reward Shaping ◽  
Multi Agent ◽  
Theoretical Results ◽  
Planning Knowledge

AbstractRecent theoretical results have justified the use of potential-based reward shaping as a way to improve the performance of multi-agent reinforcement learning (MARL). However, the question remains of how to generate a useful potential function.Previous research demonstrated the use of STRIPS operator knowledge to automatically generate a potential function for single-agent reinforcement learning. Following up on this work, we investigate the use of STRIPS planning knowledge in the context of MARL.Our results show that a potential function based on joint or individual plan knowledge can significantly improve MARL performance compared with no shaping. In addition, we investigate the limitations of individual plan knowledge as a source of reward shaping in cases where the combination of individual agent plans causes conflict.

A lateral transshipment model for perishable inventory management

2017 ◽  
Vol 55 (18) ◽  
pp. 5341-5354 ◽  
Author(s):  
Dilupa Nakandala ◽  
Henry Lau ◽  
Paul K.C. Shum
Keyword(s):  
Inventory Management ◽  
Perishable Inventory ◽  
Lateral Transshipment

scholarly journals Reinforcement Learning for Pick and Place Operations in Robotics: A Survey

Robotics ◽  
2021 ◽  
Vol 10 (3) ◽  
pp. 105
Author(s):  
Andrew Lobbezoo ◽  
Yanjun Qian ◽  
Hyock-Ju Kwon
Keyword(s):  
Reinforcement Learning ◽  
Critical Discussion ◽  
Value Iteration ◽  
Open Problems ◽  
Work Related ◽  
Model Generalization ◽  
Pick And Place ◽  
Reward Shaping ◽  
Place Task ◽  
Policy Optimization

The field of robotics has been rapidly developing in recent years, and the work related to training robotic agents with reinforcement learning has been a major focus of research. This survey reviews the application of reinforcement learning for pick-and-place operations, a task that a logistics robot can be trained to complete without support from a robotics engineer. To introduce this topic, we first review the fundamentals of reinforcement learning and various methods of policy optimization, such as value iteration and policy search. Next, factors which have an impact on the pick-and-place task, such as reward shaping, imitation learning, pose estimation, and simulation environment are examined. Following the review of the fundamentals and key factors for reinforcement learning, we present an extensive review of all methods implemented by researchers in the field to date. The strengths and weaknesses of each method from literature are discussed, and details about the contribution of each manuscript to the field are reviewed. The concluding critical discussion of the available literature, and the summary of open problems indicates that experiment validation, model generalization, and grasp pose selection are topics that require additional research.

Sign in / Sign up

Export Citation Format

Share Document