Filters








27,369 Hits in 5.1 sec

Sample-Efficient Automated Deep Reinforcement Learning [article]

Jörg K.H. Franke, Gregor Köhler, André Biedenkapp, Frank Hutter
2021 arXiv   pre-print
Despite significant progress in challenging problems across various domains, applying state-of-the-art deep reinforcement learning (RL) algorithms remains challenging due to their sensitivity to the choice  ...  In this work, we tackle the issues of sample-efficient and dynamic HPO in RL. We propose a population-based automated RL (AutoRL) framework to meta-optimize arbitrary off-policy RL algorithms.  ...  SAMPLE-EFFICIENT AUTORL In this section, we introduce a Sample-Efficient framework for Automated Reinforcement Learning (SEARL) based on an evolutionary algorithm acting on hyperparameters and gradient-based  ... 
arXiv:2009.01555v3 fatcat:ohn7ugowtrgx7erfbo36ibyhay

Curriculum-Based Deep Reinforcement Learning for Adaptive Robotics: A Mini-Review

Gupta Kashish, Najjaran Homayoun
2021 International Journal of Robotic Engineering  
This combination, Curriculum-based Deep Reinforcement Learning (CDRL), presents a powerful solution to meet the increasing complexity of today's automation industry that demands highly intelligent machines  ...  Recent progress in deep reinforcement learning has corroborated to its potential to train such autonomous and robust agents.  ...  While reinforcement learning continues to dominate the simulation-based applications, there is a principal need to study and apply sample-efficient extensions to the field, viz. curriculum learning, to  ... 
doi:10.35840/2631-5106/4131 fatcat:tnoa4vd4yrgnpjzesxr5a3jq2m

Focused section on advanced robotic systems for industrial automation

Huan Zhao, LiMin Zhu, Yunyi Jia, Naoyuki Kubota
2020 International Journal of Intelligent Robotics and Applications  
The paper "A Sample Efficient Model-based Deep Reinforcement Learning Algorithm with Experience Replay for Robot Manipulation" by Zhang et al. presents a model-based deep reinforcement learning algorithm  ...  The paper "Model accelerated reinforcement learning for high precision robotic assembly" from Zhao et al. presents a model accelerated reinforcement learning method to efficiently learn the assembly policy  ... 
doi:10.1007/s41315-020-00139-y fatcat:7gypvlwflrgszjcwkmgbsilpcq

Automating Electron Microscopy through Machine Learning and USETEM

Michael Xu, Abinash Kumar, James LeBeau
2021 Microscopy and Microanalysis  
These challenges demand a new level of automation and instrument control, crucial for efficient and reproducible electron microscopy.  ...  material-or task-specific data collection, which can be limiting given the sample heterogeneities of materials research [2] .  ...  By utilizing a combination of computer vision and deep reinforcement learning, we will show the potential for automated electron microscopy for materials science.  ... 
doi:10.1017/s1431927621010394 fatcat:6uccblcwijh73mjftxik7gx62y

Survival Analysis on Structured Data using Deep Reinforcement Learning [article]

Renith G, Harikrishna Warrier, Yogesh Gupta
2022 arXiv   pre-print
The trained model predicted larger amount of test data efficiently and performed well compared to other deep learning and machine learning models.  ...  The proposed solution involves implementation of Deep Reinforcement Learning algorithm called Double Deep Q Network (DDQN) for classifying the device failure based on the input features.  ...  Overall deep reinforcement learning algorithm can predict wide range of test data efficiently.  ... 
arXiv:2205.14331v1 fatcat:r46ofcvcofh4vgprcuxat4hane

Learning of Long-Horizon Sparse-Reward Robotic Manipulator Tasks with Base Controllers [article]

Guangming Wang, Minjian Xin, Wenhua Wu, Zhe Liu, Hesheng Wang
2021 arXiv   pre-print
Compared to previous works of learning from demonstrations, our method improves sample efficiency by orders of magnitude and improves the performance.  ...  Deep Reinforcement Learning (DRL) enables robots to perform some intelligent tasks end-to-end. However, there are still many challenges for long-horizon sparse-reward robotic manipulator tasks.  ...  deep reinforcement learning,” in 2019 International Conference on Robotics and Automation (ICRA). IEEE, 2019, pp. 754–760. [30] T. Haarnoja, A. Zhou, P. Abbeel, and S.  ... 
arXiv:2011.12105v3 fatcat:qgjdebnjmvgadkzdkn4vgqbpsm

Automated Lane Change Strategy using Proximal Policy Optimization-based Deep Reinforcement Learning [article]

Fei Ye, Xuxin Cheng, Pin Wang, Ching-Yao Chan, Jiucai Zhang
2020 arXiv   pre-print
In this study, we propose an automated lane change strategy using proximal policy optimization-based deep reinforcement learning, which shows great advantages in learning efficiency while still maintaining  ...  Machine learning-based methods offer an alternative approach, as Deep reinforcement learning (DRL) has shown promising success in many application domains including robotic manipulation, navigation, and  ...  deep reinforcement learning.  ... 
arXiv:2002.02667v2 fatcat:7beyqzig2remtnuczthjwosi7q

Accuracy-based Curriculum Learning in Deep Reinforcement Learning [article]

Pierre Fournier, Olivier Sigaud, Mohamed Chetouani, Pierre-Yves Oudeyer
2018 arXiv   pre-print
efficiency than sampling randomly.  ...  sampled randomly learns more efficiently than when asked to be very accurate at all times.  ...  ., 2018) and deep learning research (Graves et al., 2017; Matiisen et al., 2017; Florensa et al., 2017) , and are key elements of recent curiosity-driven deep reinforcement learning techniques enabling  ... 
arXiv:1806.09614v2 fatcat:wq2ukixqujbwndqhdncw3nlwxq

CYBERSECURITY INFRASTRUCTURE AND SECURITY AUTOMATION

Alex Mathew
2019 Zenodo  
AI-based security systems utilize big data and powerful machine learning algorithms to automate the security management task.  ...  The result shows that compared with the signature-based system, AI-supported security applications are efficient, accurate, and reliable.  ...  and reinforcement learning mechanisms.  ... 
doi:10.5281/zenodo.3566163 fatcat:fe6tlywwpbdgtiwijub5h2ancq

Predicting invasive ductal carcinoma using a Reinforcement Sample Learning Strategy using Deep Learning [article]

Rushabh Patel
2021 arXiv   pre-print
The aim of this study is to establish a novel method for fully automated feature extraction and classification in invasive ductal carcinoma computer-aided diagnosis (CAD) systems.  ...  The CNN layers' comprehensive structure is listed in Reinforcement sample learning strategy The aim of reinforcement sample learning is to train the model on samples that performed poorly during the  ...  Step 4: Train the CNN with normalized images using reinforcement sample learning scheme. Step 5: Use a reinforcement sample learning scheme to train the CNN with normalized images.  ... 
arXiv:2105.12564v2 fatcat:sqzaef5u2radflznjkel26adde

A Tetrahedron-Based Heat Flux Signature for Cortical Thickness Morphometry Analysis [chapter]

Yonghui Fan, Gang Wang, Natasha Lepore, Yalin Wang
2018 Lecture Notes in Computer Science  
Ultrasound Images via Layer Aggregation 281 Deep Reinforcement Learning for Vessel Centerline Tracing in Multi-modality 3D Volumes 282 Unsupervised Learning for Fast Probabilistic Diffeomorphic Registration  ...  Characterization of Hepatocellular Carcinoma 334 Deep Reinforcement Learning for Surgical Gesture Segmentation and Classification 339 Omni-supervised learning: scaling up to large unlabelled medical datasets  ... 
doi:10.1007/978-3-030-00931-1_48 pmid:30338317 pmcid:PMC6191198 fatcat:dqhvpm5xzrdqhglrfftig3qejq

Deep Reinforcement Learning framework for Autonomous Driving

AhmadEL Sallab, Mohammed Abdou, Etienne Perot, Senthil Yogamani
2017 IS&T International Symposium on Electronic Imaging Science and Technology  
Motivated by the successful demonstrations of learning of Atari games and Go by Google DeepMind, we propose a framework for autonomous driving using deep reinforcement learning.  ...  As it is a relatively new area of research for autonomous driving, we provide a short overview of deep reinforcement learning and then describe our proposed framework.  ...  Introduction Automated driving development has radically changed during the past few years, driven by advances in Artificial Intelligence (AI), and specifically Deep Learning (DL).  ... 
doi:10.2352/issn.2470-1173.2017.19.avm-023 fatcat:oarg7sq2pvay3a25i3ee43qina

Surgical Tools Detection Based on Training Sample Adaptation in Laparoscopic Videos

Guangyao Wang, Shengsheng Wang
2020 IEEE Access  
In this paper, we propose a novel framework for one-stage object detection based on a sample adaptive process controlled by reinforcement learning, which can maintain the speed advantage while maintaining  ...  The performance of object detection methods plays an important role in the recognition of surgical tools, and is a key link in the automated evaluation of surgical skills.  ...  of the long training time and low efficiency of the reinforcement learning algorithm.  ... 
doi:10.1109/access.2020.3028910 fatcat:fyl3gtcqiba7rkchc6gly3wl6m

End-to-End Automated Lane-Change Maneuvering Considering Driving Style Using a Deep Deterministic Policy Gradient Algorithm

Hongyu Hu, Ziyang Lu, Qi Wang, Chengyuan Zheng
2020 Sensors  
on reinforcement learning.  ...  Thus, autonomous vehicles trained using the proposed method can learn an automated lane-changing policy while considering safety, comfort, and efficiency.  ...  Moreover, deep reinforcement learning combines the perception ability of deep learning with the decision-making ability of reinforcement learning, thus bringing complementary advantages and providing a  ... 
doi:10.3390/s20185443 pmid:32971987 pmcid:PMC7570521 fatcat:ohgssh2xyzc4ngwnjtnigmpx7a

Classification of Material Type from Optical Coherence Tomography Images Using Deep Learning

Metin Sabuncu, Hakan Ozdemir, Stefan Wabnitz
2021 International Journal of Optics  
We present the results of the created deep learning models' capability to classify fabric fiber material types.  ...  We demonstrate that the identification procedure can be done nondestructively using optical coherence tomography (OCT) and deep learning.  ...  Reinforcement learning algorithms can make this process much quicker and smoother by enabling neural architecture search, hence automating machine learning.  ... 
doi:10.1155/2021/2520679 fatcat:dal66om52bh4lpyitooae2ze4u
« Previous Showing results 1 — 15 out of 27,369 results