检索结果-南通市图书馆

SeeMore: a spatiotemporal predictive model with bidirectional distillation and level-specific meta-adaptation

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年第8期67卷 147-171页

作者： Yuqing MA Wei LIU Yajun GAO Yang YUAN Shihao BAI Haotong QIN Xianglong LIU State key lab.of software Development Environment Beihang University

Predicting future frames using historical spatiotemporal data sequences is challenging and critical, and it is receiving a lot of attention these days from academic and industrial scholars. Most spatiotemporal predictive algorithms ignore the valuable backward reasoning ability and the disparate learning complexities among different layers and hence, cannot build good long-term dependencies and spatial correlations,resulting in suboptimal solutions. To address the aforementioned issues, we propose a two-stage coarse-to-fine spatiotemporal predictive model with bidirectional distillation and level-specific meta-adaptation(See More)in this paper, which includes a bidirectional distillation network(BDN) and a level-specific meta-adapter(LMA), to gain bidirectional multilevel reasoning. In the first stage, BDN concentrates on bidirectional dynamics modeling and coarsely constructs spatial correlations of different layers, while LMA is introduced in the second fine-tuning stage to refine the multilevel spatial correlations from a meta-learning *** particular, BDN mimics the forward and backward reasoning abilities of humans in a distillation manner,which aids in the development of long-term dependencies. The LMA views learning of different layers as disparate but related tasks and guides the transfer of learning experiences among these tasks through learning complexities. Thus, each layer could be closer to its solutions and could extract more informative spatial correlations. By capturing the enhanced short-term spatial correlations and long-term temporal dependencies,the proposed model could extract adequate knowledge from sequential historical observations and accurately predict future frames whose backtracking preconditions are consistent with the historical sequence. Our work is general and robust enough to be integrated into most spatiotemporal predictive models without requiring additional computation or memory cost during inference. Extensive experimen

关键词： spatiotemporal predictive learning knowledge transfer bidirectional distillation network level-specific meta-adapter coarse-to-fine training

Building a domain-specific compiler for emerging processors with a reusable approach

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年第1期67卷 20-38页

作者： Mingzhen LI Yi LIU Bangduo CHEN Hailong YANG Zhongzhi LUAN Depei QIAN State key lab.ratory of software Development Environment School of Computer Science and engineering Beihang University

High-performance computing and deep learning domains have been motivating the design of domain-specific *** these processors can provide promising computation capability,they are notorious for exotic programming *** improve programming productivity and fully exploit the performance potential of these processors,domain-specific compilers(DSCs) have been ***,building DSCs for emerging processors requires tremendous engineering efforts because the commonly used compilation stack is difficult to be *** to the advent of multilevel intermediate representation(MLIR),DSC developers can leverage reusable infrastructure to extend their customized functionalities without rebuilding the entire compilation *** this paper,we further demonstrate the effectiveness of MLIR by extending its reusable infrastructure to embrace a heterogeneous many-core processor(Sunway processor).In particular,we design a new Sunway dialect and corresponding backend for the Sunway processor,fully exploiting its architectural advantage and hiding its programming *** show the ease of building a DSC,we leverage the Sunway dialect and existing MLIR dialects to build a stencil compiler for the Sunway *** experimental results show that our stencil compiler,built with a reusable approach,can even perform better than state-of-the-art stencil compilers.

关键词： domain-specific compiler emerging processor reusable dialect performance optimization MLIR

ISM: intra-class similarity mixing for time series augmentation

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Frontiers of Computer Science 2024年第6期18卷 1-3页

作者： LIU Pin WANG Rui HE Yongqiang WANG Yuzhu School of Information engineering China University of Geosciences State key lab.of software Development Environment Beihang University

We propose a method based on intra-class similarity mixing, effectively alleviating the inability to improve the generalization of classification models. Its core concept involves improving the feature diversity and expanding the data scale by mixing local regions with similar features. The experimental results show that the proposed method can significantly outperform the similarity-based state-of-the-art methods. In future work, we will explore how to build and quantify the relationship between local similarity and lassification characteristics to further improve performance.

关键词： Mixing

Deep learning for code generation: a survey

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年第9期67卷 5-40页

作者： Huangzhao ZHANG Kechi ZHANG Zhuo LI Jia LI Jia LI Yongmin LI Yunfei ZHAO Yuqi ZHU Fang LIU Ge LI Zhi JIN key lab.of High Confidence software Technologies (Peking University) Ministry of Education School of Computer Science Peking University School of Computer Science and engineering Beihang University

In the past decade, thanks to the powerfulness of deep-learning techniques, we have witnessed a whole new era of automated code generation. To sort out developments, we have conducted a comprehensive review of solutions to deep learning-based code generation. In this survey, we generally formalize the pipeline and procedure of code generation and categorize existing solutions according to taxonomy from perspectives of architecture, model-agnostic enhancing strategy, metrics, and tasks. In addition, we outline the challenges faced by current dominant large models and list several plausible directions for future research. We hope that this survey may provide handy guidance to understanding, utilizing, and developing deep learning-based code-generation techniques for researchers and practitioners.

关键词： code generation automated software engineering deep learning large model artificial intelligence

Digital twin driven and intelligence enabled content delivery in end-edge-cloud collab.rative 5G networks

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Digital Communications and Networks 2024年第2期10卷 328-336页

作者： Bo Yi Jianhui Lv Xingwei Wang Lianbo Ma Min Huang College of Computer Science and engineering Northeastern UniversityShenyang110169China Pengcheng lab. Shenzhen518055China College of software Northeastern UniversityShenyang110819China College of Information Science and engineering Northeastern UniversityShenyang110819China

The rapid development of 5G/6G and AI enables an environment of Internet of Everything(IoE)which can support millions of connected mobile devices and applications to operate smoothly at high speed and low ***,these massive devices will lead to explosive traffic growth,which in turn cause great burden for the data transmission and content *** challenge can be eased by sinking some critical content from cloud to *** this case,how to determine the critical content,where to sink and how to access the content correctly and efficiently become new *** work focuses on establishing a highly efficient content delivery framework in the IoE *** particular,the IoE environment is re-constructed as an end-edge-cloud collab.rative system,in which the concept of digital twin is applied to promote the *** on the digital asset obtained by digital twin from end users,a content popularity prediction scheme is firstly proposed to decide the critical content by using the Temporal Pattern Attention(TPA)enabled Long Short-Term Memory(LSTM)***,the prediction results are input for the proposed caching scheme to decide where to sink the critical content by using the Reinforce Learning(RL)***,a collab.rative routing scheme is proposed to determine the way to access the content with the objective of minimizing *** experimental results indicate that the proposed schemes outperform the state-of-the-art benchmarks in terms of the caching hit rate,the average throughput,the successful content delivery rate and the average routing overhead.

关键词： Digital twin IoE Content delivery Caching Routing

Learnware:small models do big

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年第1期67卷 39-50页

作者： Zhi-Hua ZHOU Zhi-Hao TAN National key lab.ratory for Novel software Technology Nanjing University

There are complaints about current machine learning techniques such as the requirement of a huge amount of training data and proficient training skills,the difficulty of continual learning,the risk of catastrophic forgetting,and the leaking of data privacy/*** research efforts have been focusing on one of those concerned issues separately,paying less attention to the fact that most issues are entangled in *** prevailing big model paradigm,which has achieved impressive results in natural language processing and computer vision applications,has not yet addressed those issues,whereas becoming a serious source of carbon *** article offers an overview of the learnware paradigm,which attempts to enable users not to need to build machine learning models from scratch,with the hope of reusing small models to do things even beyond their original purposes,where the key ingredient is the specification which enables a trained model to be adequately identified to reuse according to the requirement of future users who know nothing about the model in advance.

关键词： artificial intelligence machine learning learnware

Partially-hiding functional encryption for degree-2 polynomials with fine-grained access control

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Frontiers of Computer Science 2024年第3期18卷 231-233页

作者： Haifeng QIAN Cheng LIN Qiaohan CHU Jie CHEN Shanghai key lab.ratory of Trustworthy Computing Software Engineering InstituteEast China Normal UniversityShanghai 200062China

1 Introduction Boneh et al.[1]first systematically introduce the concept of Functional Encryption(FE)which overcomes all-or-nothing limitation of traditional public key ***,FE hides a potential problem:even if the secret key is obtained legitimately,once it is released,it can be used *** leads to that the secret key may leak information *** deal with this problem and make FE more practical,Abdalla et al.[2]recently constructed inner-product FE(IPFE)with fine-grained access control achieving strong security guarantees under standard assumptions.

关键词： grain Encryption encryption

A survey on model-based reinforcement learning

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年第2期67卷 59-84页

作者： Fan-Ming LUO Tian XU Hang LAI Xiong-Hui CHEN Weinan ZHANG Yang YU National key lab.ratory for Novel software Technology Nanjing University Polixir. ai Department of Computer Science and engineering Shanghai Jiao Tong University

Reinforcement learning(RL) interacts with the environment to solve sequential decision-making problems via a trial-and-error approach. Errors are always undesirable in real-world applications, even though RL excels at playing complex video games that permit several trial-and-error attempts. To improve sample efficiency and thus reduce errors, model-based reinforcement learning(MBRL) is believed to be a promising direction, as it constructs environment models in which trial-and-errors can occur without incurring actual costs. In this survey, we investigate MBRL with a particular focus on the recent advancements in deep RL. There is a generalization error between the learned model of a non-tabular environment and the actual environment. Consequently, it is crucial to analyze the disparity between policy training in the environment model and that in the actual environment, guiding algorithm design for improved model learning, model utilization, and policy training. In addition, we discuss the recent developments of model-based techniques in other forms of RL, such as offline RL, goal-conditioned RL, multi-agent RL, and meta-RL. Furthermore,we discuss the applicability and benefits of MBRL for real-world tasks. Finally, this survey concludes with a discussion of the promising future development prospects for MBRL. We believe that MBRL has great unrealized potential and benefits in real-world applications, and we hope this survey will encourage additional research on MBRL.

关键词： reinforcement learning model-based reinforcement learning planning model learning model learning with reduced error model usage

Planning of DC Electric Spring with Particle Swarm Optimization and Elitist Non-dominated Sorting Genetic Algorithm

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

CSEE Journal of Power and Energy Systems 2024年第2期10卷 574-583页

作者： Qingsong Wang Siwei Li Hao Ding Ming Cheng Giuseppe Buja School of Electrical engineering and Jiangsu Provincial key lab.ratory of Smart Grid Technology and Equipment Southeast UniversityNanjing 210096China School of software engineering Southeast universitySuzhou 215123China

This paper addresses the planning problem of parallel DC electric springs (DCESs). DCES, a demand-side management method, realizes automatic matching of power consumption and power generation by adjusting non-critical load (NCL) and internal storage. It can offer higher power quality to critical load (CL), reduce power imbalance and relieve pressure on energy storage systems (RESs). In this paper, a planning method for parallel DCESs is proposed to maximize stability gain, economic benefits, and penetration of RESs. The planning model is a master optimization with sub-optimization to highlight the priority of objectives. Master optimization is used to improve stability of the network, and sub-optimization aims to improve economic benefit and allowable penetration of RESs. This issue is a multivariable nonlinear mixed integer problem, requiring huge calculations by using common solvers. Therefore, particle Swarm optimization (PSO) and Elitist non-dominated sorting genetic algorithm (NSGA-II) were used to solve this model. Considering uncertainty of RESs, this paper verifies effectiveness of the proposed planning method on IEEE 33-bus system based on deterministic scenarios obtained by scenario analysis.

关键词： DC distribution network DC electric spring non-dominated sorting genetic algorithm particle swarm optimization renewable energy source