Control the population of free viruses in nonlinear uncertain HIV system using Q-learning

Gholizade‐Narm, Hossein; Noori, Amin

doi:10.1007/s13042-017-0639-y

Cited by 11 publications

(2 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The authors in [11] used the Q-learning algorithm in HIV treatment and obtained a good performance and high functionality in controlling the free virions for both certain and uncertain HIV models. A mixture-of-experts approach was proposed in [2] to combine the strengths of both kernel-based regression methods (i.e., history-alignment model) and RL (i.e., model-based Bayesian POMDP model) for HIV therapy selection.…”

Section: Related Workmentioning

confidence: 99%

Incorporating causal factors into reinforcement learning for dynamic treatment regimes in HIV

Dong

Liu

et al. 2019

BMC Med Inform Decis Mak

View full text Add to dashboard Cite

Background Reinforcement learning (RL) provides a promising technique to solve complex sequential decision making problems in health care domains. However, existing studies simply apply naive RL algorithms in discovering optimal treatment strategies for a targeted problem. This kind of direct applications ignores the abundant causal relationships between treatment options and the associated outcomes that are inherent in medical domains. Methods This paper investigates how to integrate causal factors into an RL process in order to facilitate the final learning performance and increase explanations of learned strategies. A causal policy gradient algorithm is proposed and evaluated in dynamic treatment regimes (DTRs) for HIV based on a simulated computational model. Results Simulations prove the effectiveness of the proposed algorithm for designing more efficient treatment protocols in HIV, and different definitions of the causal factors could have significant influence on the final learning performance, indicating the necessity of human prior knowledge on defining a suitable causal relationships for a given problem. Conclusions More efficient and robust DTRs for HIV can be derived through incorporation of causal factors between options of anti-HIV drugs and the associated treatment outcomes.

show abstract

Section: Related Workmentioning

confidence: 99%

Incorporating causal factors into reinforcement learning for dynamic treatment regimes in HIV

Dong

Liu

et al. 2019

BMC Med Inform Decis Mak

View full text Add to dashboard Cite

show abstract

“…The cancer cells were controlled, employing Batch Reinforcement Learning (RL) method without directly adjusting the genes (Sirin, Polat, & Alhajj, 2013). By determining the optimal drug dosage, the population of the free viruses in Human Immunodeficiency Viruses (HIV) patients was controlled (Gholizade‐Narm & Noori, 2017). In this method, the eligibility traces and Q‐learning algorithms were utilized to control the free viruses, with the most optimal dosage of the drug.…”

Section: Introductionmentioning

confidence: 99%

An intelligent control strategy for cancer cells reduction in patients with chronic myelogenous leukaemia using the reinforcement learning and considering side effects of the drug

2020

Self Cite

View full text Add to dashboard Cite

Chronic Myelogenous Leukaemia (CML) is a haematopoietic stem cells disease with complex dynamical behaviour. One of the effective factors in treating patients is to determine the appropriate drug dosage. A physician should test the different drug dosages through trial and error in order to find its optimal value. This procedure is normally a time‐consuming and error‐prone task that can even be harmful. The contribution of this paper is to design an intelligent control strategy, which can be used to help physicians, by finding a drug treatment regimen to minimize the number of cancer cells for a CML patient. In this paper, the eligibility traces algorithm and Q‐learning approach are adopted as sub‐optimal methods for progressively reducing the population of cancer cells. In addition, the injected dosage of the drug has improved, compared with previous methods. More importantly, the proposed method is followed by the reduction in side effects of the drug. The advantage of the backward view and the previous states investigation are applied in the Eligibility Traces algorithm. These effects increase the learning procedure and decrease the growth rate of cancer cells and total dosage of the injected drug during the treatment period of time. The proposed strategy mitigates the side effects of the drug on the normal cells.

show abstract

Optimal control problem of various epidemic models with uncertainty based on deep reinforcement learning

Hwang

Kwon

Lee

2022

Numerical Methods Partial

View full text Add to dashboard Cite

We investigate an optimal control problem of various epidemic models with uncertainty using stochastic differential equations, random differential equations, and agent-based models. We discuss deep reinforcement learning (RL), which combines RL with deep neural networks, as one method to solve the optimal control problem. The deep Q-network algorithm is introduced to approximate an action-value function and consequently obtain the optimal policy. Numerical simulations show that in order to effectively prevent the spread of infectious diseases, it is essential to vaccinate at the highest rate for the first few days and then gradually reduce the rate.

show abstract

Control the population of free viruses in nonlinear uncertain HIV system using Q-learning

Cited by 11 publications

References 27 publications

Incorporating causal factors into reinforcement learning for dynamic treatment regimes in HIV

Incorporating causal factors into reinforcement learning for dynamic treatment regimes in HIV

An intelligent control strategy for cancer cells reduction in patients with chronic myelogenous leukaemia using the reinforcement learning and considering side effects of the drug

Optimal control problem of various epidemic models with uncertainty based on deep reinforcement learning

Contact Info

Product

Resources

About