Showing items 1 - 5 of 5

A multi-objective deep reinforcement learning framework

- Nguyen, Thanh, Nguyen, Ngoc, Vamplew, Peter, Nahavandi, Saeid, Dazeley, Richard, Lim, Chee

  • Authors: Nguyen, Thanh , Nguyen, Ngoc , Vamplew, Peter , Nahavandi, Saeid , Dazeley, Richard , Lim, Chee
  • Date: 2020
  • Type: Text , Journal article
  • Relation: Engineering Applications of Artificial Intelligence Vol. 96, no. (2020), p.
  • Full Text:
  • Reviewed:
  • Description: This paper introduces a new scalable multi-objective deep reinforcement learning (MODRL) framework based on deep Q-networks. We develop a high-performance MODRL framework that supports both single-policy and multi-policy strategies, as well as both linear and non-linear approaches to action selection. The experimental results on two benchmark problems (two-objective deep sea treasure environment and three-objective Mountain Car problem) indicate that the proposed framework is able to find the Pareto-optimal solutions effectively. The proposed framework is generic and highly modularized, which allows the integration of different deep reinforcement learning algorithms in different complex problem domains. This therefore overcomes many disadvantages involved with standard multi-objective reinforcement learning methods in the current literature. The proposed framework acts as a testbed platform that accelerates the development of MODRL for solving increasingly complicated multi-objective problems. © 2020 Elsevier Ltd

A prioritized objective actor-critic method for deep reinforcement learning

- Nguyen, Ngoc, Nguyen, Thanh, Vamplew, Peter, Dazeley, Richard, Nahavandi, Saeid

  • Authors: Nguyen, Ngoc , Nguyen, Thanh , Vamplew, Peter , Dazeley, Richard , Nahavandi, Saeid
  • Date: 2021
  • Type: Text , Journal article
  • Relation: Neural Computing and Applications Vol. 33, no. 16 (2021), p. 10335-10349
  • Full Text: false
  • Reviewed:
  • Description: An increasing number of complex problems have naturally posed significant challenges in decision-making theory and reinforcement learning practices. These problems often involve multiple conflicting reward signals that inherently cause agents’ poor exploration in seeking a specific goal. In extreme cases, the agent gets stuck in a sub-optimal solution and starts behaving harmfully. To overcome such obstacles, we introduce two actor-critic deep reinforcement learning methods, namely Multi-Critic Single Policy (MCSP) and Single Critic Multi-Policy (SCMP), which can adjust agent behaviors to efficiently achieve a designated goal by adopting a weighted-sum scalarization of different objective functions. In particular, MCSP creates a human-centric policy that corresponds to a predefined priority weight of different objectives. Whereas, SCMP is capable of generating a mixed policy based on a set of priority weights, i.e., the generated policy uses the knowledge of different policies (each policy corresponds to a priority weight) to dynamically prioritize objectives in real time. We examine our methods by using the Asynchronous Advantage Actor-Critic (A3C) algorithm to utilize the multithreading mechanism for dynamically balancing training intensity of different policies into a single network. Finally, simulation results show that MCSP and SCMP significantly outperform A3C with respect to the mean of total rewards in two complex problems: Food Collector and Seaquest. © 2021, The Author(s), under exclusive licence to Springer-Verlag London Ltd. part of Springer Nature.

A systematic review of effort-reward imbalance among health workers

- Nguyen, Huy, Le, Ma, Nguyen, Thanh, Ngoc, Dung, Ngoc, Anh, Nguyen, Phuong

  • Authors: Nguyen, Huy , Le, Ma , Nguyen, Thanh , Ngoc, Dung , Ngoc, Anh , Nguyen, Phuong
  • Date: 2018
  • Type: Text , Journal article , Review
  • Relation: International Journal of Health Planning and Management Vol. 33, no. 3 (2018), p. e674-e695
  • Full Text:
  • Reviewed:
  • Description: The purpose of this article is to systematically collate effort-reward imbalance (ERI) rates among health workers internationally and to assess gender differences. The effort-reward (ER) ratio ranges quite widely from 0.47 up to 1.32 and the ERI rate from 3.5% to 80.7%. Many studies suggested that health workers contribute more than they are rewarded, especially in Japan, Vietnam, Greece, and Germany—with ERI rates of 57.1%, 32.3%, 80.7%, and 22.8% to 27.6%, respectively. Institutions can utilize systems such as the new appraisal and reward system, which is based on performance rather than the traditional system, seniority, which creates a more competitive working climate and generates insecurity. Additionally, an increased workload and short stay patients are realities for workers in a health care environment, while the structure of human resources for health care remains inadequate. Gender differences within the ER ratio can be explained by the continued impact of traditional gender roles on attitudes and motivations that place more pressure to succeed for men rather than for women. This systematic review provides some valued evidence for public health strategies to improve the ER balance among health workers in general as well as between genders in particular. An innovative approach for managing human resources for health care is necessary to motivate and value contributions made by health workers. Copyright © 2018 John Wiley & Sons, Ltd.

Collision-free minimum-time trajectory planning for multiple vehicles based on ADMM

- Nguyen, Thanh, Nguyen, Thang, Nghiem, Truong, Nguyen, Linh, Baca, Jose, Rangel, Pablo, Song, Hyoung-Kyu

Healthcare workers' knowledge and attitudes regarding the World Health Organization's "my 5 moments for hand hygiene" : evidence from a Vietnamese central general hospital

- Nguyen, Huy, Tran, Hieu, Khuong, Long, Nguyen, Thanh, Ho, Na

  • Authors: Nguyen, Huy , Tran, Hieu , Khuong, Long , Nguyen, Thanh , Ho, Na
  • Date: 2020
  • Type: Text , Journal article
  • Relation: Journal of Preventive Medicine and Public Health Vol. 53, no. 4 (2020), p. 236-244
  • Full Text:
  • Reviewed:
  • Description: Objectives: Although the World Health Organization (WHO) initiative "My 5 Moments for Hand Hygiene"has been lauded as effective in preventing hospital-associated infections, little is known about healthcare workers (HCWs)' hand hygiene behavior. In this study, we sought to assess knowledge and attitudes towards the concepts in this initiative, as well as associated factors, among Vietnamese HCWs at a general hospital. Methods: A structured questionnaire was administered to HCWs at a central Vietnamese general hospital in 2015. Multiple logistic regression analysis was used to identify factors associated with HCWs' knowledge and attitudes towards hand hygiene. Results: Of 120 respondents, 65.8% and 67.5% demonstrated appropriate knowledge and a positive attitude, respectively, regarding all 5 hand hygiene moments. Logistic regression indicated better knowledge of hand hygiene in workers who were over 30 years old, who were direct HCWs (rather than managers), who had frequent access to clinical information, and who received their clinical information from training. Those who worked in infectious and tropical disease wards, who had frequent access to clinical information, and who received information from training were more likely to have a positive attitude towards hand hygiene than their counterparts. Conclusions: Although many Vietnamese HCWs displayed moderate knowledge and positive attitudes towards the WHO hand hygiene guidelines, a key gap remained. Regular education and training programs are needed to increase knowledge and to improve attitudes and practices towards hand hygiene. Furthermore, a combination of multimodal strategies and locally-adapted interventions is needed for sustainable hand hygiene adherence. Copyright © 2020 The Korean Society for Preventive Medicine. **Please note that there are multiple authors for this article therefore only the name of the first 5 including Federation University Australia affiliate “Huy Nguyen” is provided in this record**
  • 1