Small-sample reinforcement learning: Improving policies using synthetic data1

Carden, Stephen W.; Livsey, James

doi:10.3233/idt-170285

Cited by 3 publications

(3 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Mahmud et al [ 20 ] and Zhao [ 21 ] proposed a fault diagnosis method based on variational AE and convolutional neural network to solve the problems of few fault samples and imbalanced data in the fault diagnosis method of the above-mentioned drive. Carden [ 22 ] proposed a strategy learning method, which uses part of the researchers' knowledge of probability transfer structure to transform it into an approximate generation model, from which to generate synthetic data.…”

Section: Related Workmentioning

confidence: 99%

Product Processing Quality Classification Model for Small-Sample and Imbalanced Data Environment

Liu

Dai

2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

With the rapid development of machine learning technology, how to use machine learning technology to empower the manufacturing industry has become a research hotspot. In order to solve the problem of product quality classification in a small sample data and imbalanced data environment, this paper proposes a data generation model called MSMOTE-GAN, which is based on Mahalanobis Synthetic Minority Oversampling Technology (MSMOTE) and Generative Adversarial Network (GAN). Among them, MSMOTE is proposed to solve the problem of the sample biased to the majority class expanded by methods such as GAN in a sample imbalanced environment. Based on the traditional SMOTE method, the sample distance measurement method is modified from Euclidean distance to Mahalanobis distance, taking into account the correlation between attributes and the influence of dimensions on the sample distance. In the data generation model, MSMOTE is used to balance the positive and negative samples in the data. GAN generates fake data with the same distribution as the original data based on a balanced data set and expands the sample size to solve the problems of overfitting and insufficient model expression ability that occur when the sample size is too small. The quality classification framework of water heater liner based on the data generation model and Random Forest is constructed, and the process of the quality classification of water heater liner under the environment of small sample data and imbalanced data is fully described. This paper compares the MSMOTE-GAN model, Bootstrap, and tableGAN on the water heater liner production line data set and the public data set. The experimental result shows that the expanded data set of the MSMOTE-GAN model can effectively improve the performance of the classification model.

show abstract

Section: Related Workmentioning

confidence: 99%

Product Processing Quality Classification Model for Small-Sample and Imbalanced Data Environment

Liu

Dai

2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

show abstract

“…Broadly, there are two different ways of dealing with a small sample learning problem [344]. The direct solution can be using data augmentation strategies such as deformations [345] or GANs [346] to increase samples and then employ conventional learning methods.…”

Section: Learning From Small Datamentioning

confidence: 99%

“…The other type of solutions can be applying various model modification or domain adaptation methods such as knowledge distillation [347] or meta-learning [348] to enable efficient learning that overcomes the problem of data scarcity. While still in its early stage, significant progress has been made in small sample learning research in recent years [344]. How to build on these achievements and tackle the small data RL problems in healthcare domains thus calls for new methods of future investigations.…”

Section: Learning From Small Datamentioning

confidence: 99%

Reinforcement Learning in Healthcare: A Survey

Yu¹,

Liu²,

Nemati³

2019

Preprint

View full text Add to dashboard Cite

As a subfield of machine learning, reinforcement learning (RL) aims at empowering one's capabilities in behavioural decision making by using interaction experience with the world and an evaluative feedback. Unlike traditional supervised learning methods that usually rely on one-shot, exhaustive and supervised reward signals, RL tackles with sequential decision making problems with sampled, evaluative and delayed feedback simultaneously. Such distinctive features make RL technique a suitable candidate for developing powerful solutions in a variety of healthcare domains, where diagnosing decisions or treatment regimes are usually characterized by a prolonged and sequential procedure. This survey will discuss the broad applications of RL techniques in healthcare domains, in order to provide the research community with systematic understanding of theoretical foundations, enabling methods and techniques, existing challenges, and new insights of this emerging paradigm. By first briefly examining theoretical foundations and key techniques in RL research from efficient and representational directions, we then provide an overview of RL applications in a variety of healthcare domains, ranging from dynamic treatment regimes in chronic diseases and critical care, automated medical diagnosis from both unstructured and structured clinical data, as well as many other control or scheduling domains that have infiltrated many aspects of a healthcare system. Finally, we summarize the challenges and open issues in current research, and point out some potential solutions and directions for future research.

show abstract

Reinforcement Learning in Healthcare: A Survey

et al. 2021

View full text Add to dashboard Cite

As a subfield of machine learning, reinforcement learning (RL) aims at optimizing decision making by using interaction samples of an agent with its environment and the potentially delayed feedbacks. In contrast to traditional supervised learning that typically relies on one-shot, exhaustive, and supervised reward signals, RL tackles sequential decision-making problems with sampled, evaluative, and delayed feedbacks simultaneously. Such a distinctive feature makes RL techniques a suitable candidate for developing powerful solutions in various healthcare domains, where diagnosing decisions or treatment regimes are usually characterized by a prolonged period with delayed feedbacks. By first briefly examining theoretical foundations and key methods in RL research, this survey provides an extensive overview of RL applications in a variety of healthcare domains, ranging from dynamic treatment regimes in chronic diseases and critical care, automated medical diagnosis, and many other control or scheduling problems that have infiltrated every aspect of the healthcare system. In addition, we discuss the challenges and open issues in the current research and highlight some potential solutions and directions for future research.

show abstract

Small-sample reinforcement learning: Improving policies using synthetic data1

Cited by 3 publications

References 11 publications

Product Processing Quality Classification Model for Small-Sample and Imbalanced Data Environment

Product Processing Quality Classification Model for Small-Sample and Imbalanced Data Environment

Reinforcement Learning in Healthcare: A Survey

Reinforcement Learning in Healthcare: A Survey

Contact Info

Product

Resources

About