检索结果-内蒙古大学图书馆

Deep Reinforcement Learning based Indoor Air Quality Sensing by Cooperative Mobile Robots

学校读者我要写书评

暂无评论

Deep Reinforcement Learning based Indoor Air Quality Sensing...

IEEE Conference on Wireless Communications and Networking

作者： Zhiwen Hu Tiankuo Song Kaigui Bian Lingyang Song National Engineering Laboratory for Big Data Analysis and Applications Peking University Beijing 101 Middle School Beijing China Department of Computer Science Peking University China

ISBN: (数字)9781728131061

ISBN: (纸本)9781728131078

Confronted with the severe indoor air pollution nowadays, we propose the usage of multiple robots to detect the indoor air quality (IAQ) cooperatively for fewer sensors and larger sensing area. To acquire the complete real-time IAQ distribution map, we exploit the real statistical data to construct the IAQ data model and adopt Kalman Filter to obtain the estimation of the unmeasured area. Since the movement of the robots affects the estimation accuracy, a proper movement strategy should be planned to minimize the total estimation error. To solve this optimization problem, we design a deep Q-learning approach, which provides sub-optimal movement strategies for real-time robot sensing. By simulations, we verify the adopted IAQ data model and testify the effectiveness of the proposed solution. For application considerations, we have deployed this system in peking university since Dec. 2018 and developed a website to visualize the IAQ distribution.

关键词：

A Heuristic Method for Route Programming in Puzzle-Based Energy Storage Systems

学校读者我要写书评

暂无评论

A Heuristic Method for Route Programming in Puzzle-Based Ene...

IEEE Conference on Energy Internet and Energy System Integration (EI2)

作者： Yang Zou Jianxiao Wang School of Engineering Peking University Beijing China National Engineering Laboratory for Big Data Analysis and Applications Peking University Beijing China Peking University Ordos Research Institute of Energy Ordos China

Recently, many energy storage-related enterprises have been facing difficulties brought out by the limitation of land and the increase in loan cost. As a promising approach to improving space utilization rate, puzzle-based energy storage systems (PBESSs) are drawing more attention from logistics researchers. In previous research about PBESS, concentration has been paid to single target item problems. However, there are no consensus algorithms to solve load retrievals route programming in PBESSs with multiple target items. In this paper, a heuristic algorithm is proposed to solve load retrieval route programming in PBESSs with multiple target items, multiple escorts, and multiple (Inputs Outputs) IOs. This paper defines new concepts about the proposed algorithm, including target IOs, target position of escorts, number of escorts required, et al. Then, the decision procedures are designed according to these concepts. Based on the Markov decision process, the proposed algorithm makes the decision of the next action by analyzing the current status, until all the target items arrive at the IOs. The case study results have shown the effectiveness and efficiency of the proposed heuristic algorithm.

关键词：

HET: Scaling out huge embedding model training via cache-enabled distributed framework

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Miao, Xupeng Zhang, Hailin Shi, Yining Nie, Xiaonan Yang, Zhi Tao, Yangyu Cui, Bin Peking University China Institute of Computational Social Science Peking University Qingdao China Tencent Inc Center for Data Science Peking University National Engineering Laboratory for Big Data Analysis and Applications China

Embedding models have been an effective learning paradigm for high-dimensional data. However, one open issue of embedding models is that their representations (latent factors) often result in large parameter space. We observe that existing distributed training frameworks face a scalability issue of embedding models since updating and retrieving the shared embedding parameters from servers usually dominates the training cycle. In this paper, we propose HET, a new system framework that significantly improves the scalability of huge embedding model training. We embrace skewed popularity distributions of embeddings as a performance opportunity and leverage it to address the communication bottleneck with an embedding cache. To ensure consistency across the caches, we incorporate a new consistency model into HET design, which provides fine-grained consistency guarantees on a per-embedding basis. Compared to previous work that only allows staleness for read operations, HET also utilizes staleness for write operations. Evaluations on six representative tasks show that HET achieves up to 88% embedding communication reductions and up to 20.68× performance speedup over the state-of-the-art baselines. Copyright © 2021, The Authors. All rights reserved.

关键词： Embeddings

A Novel Control Algorithm for Interaction Between Surface Waves and A Permeable Floating Structure

学校读者我要写书评

暂无评论

China Ocean engineering 2016年第2期30卷 161-176页

作者： Pei-Wei TSAI A.ALSAEDI T.HAYAT Cheng-Wu CHEN College of information Science and Engineering Fujian University of TechnologyFuzhou 350118China Fujian Provincial Key Laboratory of Big Data Mining and Applications(Fujian University of Technology) Fuzhou 350118China Nonlinear Analysis and Applied Mathematics(NAAM)Research Group King Abdulaziz UniversityJeddah 21589Saudi Arabia Department of Mathematics Quaid-I-Azam UniversityIslamabad 44000Pakistan Department of Maritime Information and Technology National Kaohsiung Marine UniversityKaohsiung 80543China Faculty of Engineering King Abdulaziz UniversityJeddah 21589Saudi Arabia

An analytical solution is undertaken to describe the wave-induced flow field and the surge motion of a permeable platform structure with fuzzy controllers in an oceanic *** the design procedure of the controller,a parallel distributed compensation（PDC） scheme is utilized to construct a global fuzzy logic controller by blending all local state feedback controllers.A stability analysis is carried out for a real structure system by using Lyapunov *** corresponding boundary value problems are then incorporated into scattering and radiation *** are analytically solved,based on separation of variables,to obtain series solutions in terms of the harmonic incident wave motion and surge *** dependence of the wave-induced flow field and its resonant frequency on wave characteristics and structure properties including platform width,thickness and mass has been thus drawn with a parametric *** which mathematical models are applied for the wave-induced displacement of the surge motion.A nonlinearly inverted pendulum system is employed to demonstrate that the controller tuned by swarm intelligence method can not only stabilize the nonlinear system,but has the robustness against external disturbance.

关键词： fuzzy Lyapunov function Takagi?Sugeno form swarm intelligence algorithm

Efficient Diversity-Driven Ensemble for Deep Neural Networks

学校读者我要写书评

暂无评论

Efficient Diversity-Driven Ensemble for Deep Neural Networks

International Conference on data engineering

作者： Wentao Zhang Jiawei Jiang Yingxia Shao Bin Cui Center for Data Science Peking University & National Engineering Laboratory for Big Data Analysis and Applications ETH Zurich Switzerland School of Computer Science Beijing University of Posts and Telecommunications

ISBN: (数字)9781728129037

ISBN: (纸本)9781728129044

The ensemble of deep neural networks has been shown, both theoretically and empirically, to improve generalization accuracy on the unseen test set. However, the high training cost hinders its efficiency since we need a sufficient number of base models and each one in the ensemble has to be separately trained. Lots of methods are proposed to tackle this problem, and most of them are based on the feature that a pre-trained network can transfer its knowledge to the next base model and then accelerate the training process. However, these methods suffer a severe problem that all of them transfer knowledge without selection and thus lead to low diversity. As the effect of ensemble learning is more pronounced if ensemble members are accurate and diverse, we propose a method named Efficient Diversity-Driven Ensemble (EDDE) to address both the diversity and the efficiency of an ensemble. To accelerate the training process, we propose a novel knowledge transfer method which can selectively transfer the previous generic knowledge. To enhance diversity, we first propose a new diversity measure, then use it to define a diversity-driven loss function for optimization. At last, we adopt a Boosting-based framework to combine the above operations, such a method can also further improve diversity. We evaluate EDDE on Computer Vision (CV) and Natural Language Processing (NLP) tasks. Compared with other well-known ensemble methods, EDDE can get highest ensemble accuracy with the lowest training cost, which means it is efficient in the ensemble of neural networks.

关键词： Erbium Kuiper belt Conferences data engineering

Activation Control of Multiple Piecewise Linear Neural Networks

学校读者我要写书评

暂无评论

Activation Control of Multiple Piecewise Linear Neural Netwo...

IEEE International Conference on Automation Science and engineering (CASE)

作者： Chen Hou National Engineering Laboratory for Big Data Analysis and Applications Peking University Beijing China APKU-Changsha Institute for Computing and Digital Economy Changsha China

ISBN: (数字)9798350358513

ISBN: (纸本)9798350358520

Piecewise linear neural networks (PLNNs) are proven universal approximators for continuous functions on the compact domain. For multiple PLNNs (mPLNNs) differing from each other in suffering different approximation errors (AEs) when approximating the same continuous function, activating all the PLNNs to approximate the continuous function, and then, picking up the minimum AE (MAE) from all the AEs seems to be a practical way to arrive at such MAE. Activating PLNNs has to consume energy, and more activated PLNNs provide more AEs to consider, which can maximize the probability of guaranteeing the MAE, while also needing to harvest more energy. Therefore, how to make the optimal tradeoff between energy harvested and approximation accuracy for mPLNNs arises as an interesting issue. To address this problem, this paper first deduces the objective function, with the accumulative probability of obtaining the MAE as the objective and the accumulative energy harvested as the constraint, then proposes an algorithm for mPLNNs to enjoy the maximum probability of achieving the MAE within the acceptable level of accumulative energy harvested. Experiments verify its performance.

关键词： Energy consumption Accuracy Computer aided software engineering Automation Neurons Approximation error Approximation algorithms Linear programming Biological neural networks

Well-balanced fifth-order finite volume WENO schemes with constant subtraction technique for shallow water equations

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zhao, Lidan Tao, Zhanjing Zhang, Min School of Mathematics Jilin University Changchun130012 China National Engineering Laboratory for Big Data Analysis and Applications Peking University Beijing100871 China Chongqing Research Institute of Big Data Peking University Chongqing401121 China

In this paper, we propose a new well-balanced fifth-order finite volume WENO method for solving one- and two-dimensional shallow water equations with bottom topography. The well-balanced property is crucial to the ability of a scheme to simulate perturbation waves over the "lake-at-rest" steady state such as waves on a lake or tsunami waves in the deep ocean. We adopt the constant subtraction technique such that both the flux gradient and source term in the new pre-balanced form vanish at the lake-at-rest steady state, while the well-balanced WENO method by Xing and Shu [Commun. Comput. Phys., 2006] uses high-order accurate numerical discretization of the source term and makes the exact balance between the source term and the flux gradient, to achieve the well-balanced property. The scaling positivity-preserving limiter is used for the water height near the dry areas. The fifth-order WENO-AO reconstruction is used to construct the solution since it has better resolution than the WENO-ZQ and WENO-MR reconstructions for the perturbation of steady state flows. Extensive one- and two-dimensional numerical examples are presented to demonstrate the well-balanced, fifth-order accuracy, non-oscillatory, and positivity-preserving properties of the proposed method. Copyright © 2024, The Authors. All rights reserved.

关键词： Finite volume method

Unsupervised domain adaptation in semantic segmentation based on pixel alignment and self-training

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Dong, Hexin Yu, Fei Zhao, Jie Dong, Bin Zhang, Li Center for Data Science Peking University Beijing China Center for Data Science in Health and Medicine Peking University Beijing China National Engineering Laboratory for Big Data Analysis and Applications Peking University Beijing China Peking University Beijing China

This paper proposes an unsupervised cross-modality domain adaptation approach based on pixel alignment and self-training. Pixel alignment transfers ceT1 scans to hrT2 modality, helping to reduce domain shift in the training segmentation model. Self-training adapts the decision boundary of the segmentation network to fit the distribution of hrT2 scans. Experiment results show that PAST has outperformed the non-UDA baseline significantly, and it received rank-2 on CrossMoDA validation phase Leaderboard with a mean Dice score of 0.8395. Copyright © 2021, The Authors. All rights reserved.

关键词： Pixels

INFORMATION GAIN PROPAGATION: A NEW WAY TO GRAPH ACTIVE LEARNING WITH SOFT LABELS

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Zhang, Wentao Wang, Yexin You, Zhenbang Cao, Meng Huang, Ping Shan, Jiulong Yang, Zhi Cui, Bin School of CS Peking University China Apple United States National Engineering Laboratory for Big Data Analysis and Applications Institute of Computational Social Science Peking University Qingdao China

Graph Neural Networks (GNNs) have achieved great success in various tasks, but their performance highly relies on a large number of labeled nodes, which typically requires considerable human effort. GNN-based Active Learning (AL) methods are proposed to improve the labeling efficiency by selecting the most valuable nodes to label. Existing methods assume an oracle can correctly categorize all the selected nodes and thus just focus on the node selection. However, such an exact labeling task is costly, especially when the categorization is out of the domain of individual expert (oracle). The paper goes further, presenting a soft-label approach to AL on GNNs. Our key innovations are: i) relaxed queries where a domain expert (oracle) only judges the correctness of the predicted labels (a binary question) rather than identifying the exact class (a multi-class question), and ii) new criteria of maximizing information gain propagation for active learner with relaxed queries and soft labels. Empirical studies on public datasets demonstrate that our method significantly outperforms the state-of-the-art GNN-based AL methods in terms of both accuracy and labeling cost. Copyright © 2022, The Authors. All rights reserved.

关键词： Graph neural networks