检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Budan, Gokhan Damiani, Francesca Kurtulus, Can Ure, Nazim Kemal Eatron Technologies Ltd. Research and Development Warwick CV34 6UW United Kingdom Istanbul Technical University Artificial Intelligence and Data Science Application Research Center Istanbul Turkey

Model identification of battery dynamics is a central problem in energy research;many energy management systems and design processes rely on accurate battery models for efficiency optimization. The standard methodology for battery modelling is traditional design of experiments (DoE), where the battery dynamics are excited with many different current profiles and the measured outputs are used to estimate the system dynamics. However, although it is possible to obtain useful models with the traditional approach, the process is time consuming and expensive because of the need to sweep many different current-profile configurations. In the present work, a novel DoE approach is developed based on deep reinforcement learning, which alters the configuration of the experiments on the fly based on the statistics of past experiments. Instead of sticking to a library of predefined current profiles, the proposed approach modifies the current profiles dynamically by updating the output space covered by past measurements, hence only the current profiles that are informative for future experiments are applied. Simulations and real experiments are used to show that the proposed approach gives models that are as accurate as those obtained with traditional DoE but by using 85% less resources. © 2023, CC BY-NC-ND.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Modified Query Expansion Through Generative Adversarial Networks for Information Extraction in E-Commerce

SSRN

引用

SSRN 2023年

作者： Cakir, Altan Gurkan, Mert Physics Engineering Faculty of Science and Letters Istanbul Technical University Istanbul34469 Turkey Istanbul Technical University Artificial Intelligence Data Science Research and Application Center Istanbul34469 Turkey Istanbul34415 Turkey

This work addresses an alternative approach for query expansion (QE) using a generative adversarial network (GAN) to enhance the effectiveness of information search in e-commerce. We propose a modified QE conditional GAN (mQE-CGAN) framework, which resolves keywords by expanding the query with a synthetically generated query that proposes semantic information from text input. We train a sequence-to-sequence transformer model as the generator to produce keywords and use a recurrent neural network model as the discriminator to classify an adversarial output with the generator. With the modified CGAN framework, various forms of semantic insights gathered from the query-document corpus are introduced to the generation process. We leverage these insights as conditions for the generator model and discuss their effectiveness for the query expansion task. Our experiments demonstrate that the utilization of condition structures within the mQE-CGAN framework can increase the semantic similarity between generated sequences and reference documents up to nearly 10\% compared to baseline models. © 2023, The Authors. All rights reserved.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

ExpoNAS: Using Exposure-based Candidate Exclusion to Reduce NAS Space for Heterogeneous Pipeline of CNNs 29

ExpoNAS: Using Exposure-based Candidate Exclusion to Reduce ...

引用

29th IEEE International Conference on Parallel and Distributed systems, ICPADS 2023

作者： Duan, Huijuan Wang, Danghui Zhao, Xingting Mei, Kuizhi Wang, Jihe Polytechnical University School of Computer Science Northwestern Xi'an China Northwestern Polytechnical University Engineering Research Center of Embedded System Integration Ministry of Education Xi'an China Xi'An Jiaotong University Institute of Artificial Intelligence and Robotics Xi'an China Northwestern Polytechnical University Natl. Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology Xi'an China

ISBN: (纸本)9798350330717

Efficiently deploying neural network models on heterogeneous edge systems faces unique challenges. Two major issues are underutilized pipelines caused by hardware heterogeneity and explosive growth of architecture search space. To address these challenges, we propose ExpoNAS, a framework enabling fast deployment-oriented neural architecture search on heterogeneous platforms. ExpoNAS evaluates model partitioning and deployability before search to meet accuracy and latency constraints. However, repeated partitioning leads to exponential growth in architecture search space. We first observe a strong correlation between operational exposure and accuracy improvement, and then we propose an exposure-based reduction technique that excludes the operations of low exposure. By removing these non-critical operations, the exposure-based pruning successfully reduces architecture search time. Our experiments on popular datasets show that the ExpoNAS pipelined design improves execution efficiency by a factor of 2 compared to traditional batch processing. Without compromising accuracy, ExpoNAS is able to reduce the architecture search time by 47.5% for discovering high-accuracy networks, compared to the SOTA NAS strategies. © 2023 IEEE.

关键词： heterogeneous systems NAS pipeline mapping search space reduction

来源：评论

学校读者我要写书评

暂无评论

Enhanced prairie dog optimization with Levy flight and dynamic opposition-based learning for global optimization and engineering design problems

引用

Neural Computing and applications 2024年第19期36卷 11137-11170页

作者： Biswas, Saptadeep Shaikh, Azharuddin Ezugwu, Absalom El-Shamir Greeff, Japie Mirjalili, Seyedali Bera, Uttam Kumar Abualigah, Laith Department of Mathematics National Institute of Technology Agartala Tripura Agartala India Unit for Data Science and Computing North-West University 11 Hoffman Street Potchefstroom2520 South Africa Faculty of Natural and Agricultural Sciences School of Computer Science and Information Systems North-West University Vanderbijlpark South Africa Centre for Artificial Intelligence Research and Optimization Torrens University Australia Fortitude Valley BrisbaneQLD4006 Australia Research Center University of Tabuk Tabuk71491 Saudi Arabia Computer Science Department Al al-Bayt University Mafraq25113 Jordan Hourani Center for Applied Scientific Research Al-Ahliyya Amman University Amman19328 Jordan Applied Science Research Center Applied Science Private University Amman11931 Jordan Department of Electrical and Computer Engineering Lebanese American University Byblos13-5053 Lebanon School of Engineering and Technology Sunway University Malaysia Petaling Jaya27500 Malaysia College of Engineering Yuan Ze University Taoyuan32003 Taiwan MEU Research Unit Middle East University Amman11831 Jordan

This study proposes a new prairie dog optimization algorithm version called EPDO. This new version aims to address the issues of premature convergence and slow convergence that were observed in the original PDO algorithm. To improve performance, several modifications are introduced in EPDO. First, a dynamic opposite learning strategy is employed to increase the diversity of the population and prevent premature convergence. This strategy helps the algorithm avoid falling into local optima and promotes global optimization. Additionally, the Lévy dynamic random walk technique is utilized in EPDO. This modified Lévy flight with random walk reduces the algorithm’s running time for the test function’s ideal value, accelerating its convergence. The proposed approach is evaluated using 33 benchmark problems from CEC 2017 and compared against seven other comparative techniques: GWO, MFO, ALO, WOA, DA, SCA, and RSA. Numerical results demonstrate that EPDO produces good outcomes and performs well in solving benchmark problems. To further validate the results and assess reliability, the authors employ average rank tests, the measurement of alternatives, and ranking according to the compromise solution (MARCOS) method, as well as a convergence report of EPDO and other algorithms. Furthermore, the effectiveness of the EPDO algorithm is demonstrated by applying it to five design problems. The results indicate that EPDO achieves impressive outcomes and proves its capability to address practical issues. The algorithm performs well in solving benchmark and practical design problems, as supported by the numerical results and validation methods used in the study. © The Author(s) 2024.

关键词： Global optimization

来源：评论

学校读者我要写书评

暂无评论

Omni-Granularity Embedding Network for Text-to-Image Person Retrieval

Omni-Granularity Embedding Network for Text-to-Image Person ...

引用

IEEE International Conference on Multimedia and Expo (ICME)

作者： Chengji Wang Zhiming Luo Shaozi Li Hubei Provincial Key Laboratory of Artificial Intelligence and Smart Learning National Language Resources Monitoring and Research Center for Network Media School of Computer Science Central China Normal University Wuhan China Department of Artificial Intelligence Xiamen University Xiamen China Fujian Key Laboratory of Big Data Application and Intellectualization for Tea Industry WuyiShan China

ISBN: (数字)9798350390155

ISBN: (纸本)9798350390162

Text-to-image person retrieval aims to identify the desired individual based on a textual description. As an instance-level retrieval problem, it has a large intra-class variance and a small inter-class variance. Although significant progress has been made, the omni-granularity matching issue remains unaddressed. Omni-granularity matching involves aligning words with multi-granularity image regions, challenging models to learn in an omni-granularity embedding space. In this paper, we introduce a novel Omni-Granularity Embedding Network (OGEN) for person representation learning. It addresses the omni-granularity matching issue by developing a Cross-Granularity Aggregation Module (CGAM). This module dynamically consolidates diverse granularity features for learning granularity-dependent and omni-granularity person representations. Additionally, a teacher-student knowledge transfer framework is introduced to minimize the inter-modality discrepancy, allowing CGAM to focus on modality-shared semantics. Due to the effectiveness of CGAM and the knowledge transfer framework, our OGEN enhances the Rank-1 accuracy of the Baseline by 8.54%, 9.89%, and 11.09% on three public datasets, respectively.

关键词： Representation learning Accuracy Semantics Text to image Knowledge transfer

来源：评论

学校读者我要写书评

暂无评论

An Efficient Model-Agnostic Approach for Uncertainty Estimation in data-Restricted Pedometric applications

An Efficient Model-Agnostic Approach for Uncertainty Estimat...

引用

International Conference on Machine Learning and applications (ICMLA)

作者： Viacheslav Barkov Jonas Schmidinger Robin Gebbers Martin Atzmueller Joint Lab Artificial Intelligence and Data Science Osnabrück University Osnabrück Germany Department of Agromechatronics Leibniz Institute for Agricultural Engineering and Bioeconomy Potsdam Germany Semantic Information Systems Group Osnabrück University Osnabrück Germany German Research Center for Artificial Intelligence (DFKI) Osnabrück Germany

ISBN: (数字)9798350374889

ISBN: (纸本)9798350374896

This paper introduces a model-agnostic approach designed to enhance uncertainty estimation in the predictive modeling of soil properties, a crucial factor for advancing pedometrics and the practice of digital soil mapping. For ad-dressing the typical challenge of data scarcity in soil studies, we present an improved technique for uncertainty estimation. This method is based on the transformation of regression tasks into classification problems, which not only allows for the production of reliable uncertainty estimates but also enables the application of established machine learning algorithms with competitive performance that have not yet been utilized in pedometrics. Empirical results from datasets collected from two German agricultural fields showcase the practical application of the proposed methodology. Our results and findings suggest that the proposed approach has the potential to provide better uncertainty estimation than the models commonly used in pedometrics.

关键词： Analytical models Uncertainty Machine learning algorithms Estimation Transforms Soil Predictive models Transformers data models Reliability

来源：评论

学校读者我要写书评

暂无评论

Self-Improving Safety Performance of Reinforcement Learning Based Driving with Black-Box Verification Algorithms

arXiv

引用

arXiv 2022年

作者： Dagdanov, Resul Durmus, Halil Ure, Nazim Kemal ITU Artificial Intelligence and Data Science Research Center The Department of Aeronautical Engineering Istanbul Technical University Turkey Eatron Technologies The Department of Electronics and Communication Engineering Istanbul Technical University Turkey ITU Artificial Intelligence and Data Science Application and Research Center Department of Computer Engineering Istanbul Technical University Turkey

In this work, we propose a self-improving artificial intelligence system to enhance the safety performance of reinforcement learning (RL)-based autonomous driving (AD) agents using black-box verification methods. RL algorithms have become popular in AD applications in recent years. However, the performance of existing RL algorithms heavily depends on the diversity of training scenarios. A lack of safety-critical scenarios during the training phase could result in poor generalization performance in real-world driving applications. We propose a novel framework in which the weaknesses of the training set are explored through black-box verification methods. After discovering AD failure scenarios, the RL agent’s training is re-initiated via transfer learning to improve the performance of previously unsafe scenarios. Simulation results demonstrate that our approach efficiently discovers safety failures of action decisions in RL-based adaptive cruise control (ACC) applications and significantly reduces the number of vehicle collisions through iterative applications of our method. The source code is publicly available at https://***/data-and-decision-lab/self-improving-RL. © 2022, CC BY.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Transferring Traffic Predictions to Urban Regions Without Target data

Transferring Traffic Predictions to Urban Regions Without Ta...

引用

International Conference on Intelligent Transportation

作者： Stefan Schestakov Simon Gottschalk Nicolas Tempelmeier Thorben Funke Elena Demidova L3S Research Center Leibniz University Hannover Hannover Germany Volkswagen AG Commercial Vehicles Hannover Germany Data Science & Intelligent Systems (DSIS) Research Group University of Bonn and Lamarr Institute for Machine Learning and Artificial Intelligence Bonn Germany

ISBN: (数字)9798331505929

ISBN: (纸本)9798331505936

The scarcity of spatiotemporal traffic data for many urban regions significantly limits the availability of location-specific predictive models for traffic management, mobility services, and road safety. For example, whereas some cities release taxi data for several years, for many other cities traffic data is not available. Although existing transfer learning approaches aim to alleviate this problem, they require at least some data to be available in the target urban region. Traffic prediction in the regions without data remains a critical and unsolved task. To this extent, we propose Zero-ST, a novel transfer learning approach that leverages inductive road features and road topology to learn latent road representations encoding traffic pattern similarity. Our experimental evaluation on real-world data demonstrates the effectiveness of Zero-ST for the traffic speed prediction task in several urban regions. On this task, Zero-ST achieves an error reduction of at least 4.77% concerning MAE compared to the baselines.

关键词： Roads Transfer learning Urban areas Traffic control Predictive models Road safety Spatiotemporal phenomena Topology Public transportation Knowledge transfer

来源：评论

学校读者我要写书评

暂无评论

Comparison of spin-correlation and polarization variables of spin density matrix for top quark pairs at the LHC and New Physics implications

arXiv

引用

arXiv 2022年

作者： Cakir, Altan Kolay, Orcun Department of Physics Engineering Istanbul Technical University Maslak Istanbul Turkey Istanbul Technical University Artificial Intelligence Data Science Research and Application Center Maslak Istanbul Turkey

Precise determination of top-quark pairs is an essential tool for understanding the overall consistency of the standard model (SM) expectations, understanding limited New Physics (NP) models, through spin-spin correlation and polarization parameters, and has a critical impact on the analyses strategies at upcoming LHC programs. In this work, we review and discuss various state-of-the-art Monte Carlo (MC) methodologies as MadGraph5_aMC@NLO, Sherpa, Powheg-Box and Pythia8, which are Matrix Element (ME)/Parton Shower (PS) matching generators including a complete set of spin correlation and polarization in top quark pair production with dileptonic final states. This is the first such study that not only compares the effects of different MC event generator approaches on spin density matrix elements and polarization parameters, but also investigates the effects of leading order (LO) and next-to-leading order (NLO) accuracy in QCD, and electroweak (EW) corrections via Sherpa. Moreover, as a continuation of the work, the prospects for possible NP scenarios through top-quark spin-spin correlation and polarization measurements for Supersymmetry (R parity conserved and violated models) and Dark Matter (top quarks associated mediator) models during upcoming LHC runs are briefly outlined. We find that all SM MC predictions for the defined set of variables are generally consistent with the experimental data and theoretical predictions within the uncertainty variations. Besides, for the distributions of the cos , laboratory-frame observables (cos lab and |∆φll|) and the observables generated by parity (P) and charge-parity (CP) conserving interactions, we conclude some clues that the considered signals and beyond may well be separated from experimental data and located above the SM predictions. © 2022, CC BY.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Modified Query Expansion Through Generative Adversarial Networks for Information Extraction in E-Commerce

arXiv

引用

arXiv 2022年

作者： Cakir, Altan Gurkan, Mert Physics Engineering Faculty of Science and Letters Istanbul Technical University Istanbul Turkey Istanbul Technical University Artificial Intelligence Data Science Research and Application Center Istanbul Turkey Istanbul Turkey

This work addresses an alternative approach for query expansion (QE) using a generative adversarial network (GAN) to enhance the effectiveness of information search in e-commerce. We propose a modified QE conditional GAN (mQE-CGAN) framework, which resolves keywords by expanding the query with a synthetically generated query that proposes semantic information from text input. We train a sequence-to-sequence transformer model as the generator to produce keywords and use a recurrent neural network model as the discriminator to classify an adversarial output with the generator. With the modified CGAN framework, various forms of semantic insights gathered from the query-document corpus are introduced to the generation process. We leverage these insights as conditions for the generator model and discuss their effectiveness for the query expansion task. Our experiments demonstrate that the utilization of condition structures within the mQE-CGAN framework can increase the semantic similarity between generated sequences and reference documents up to nearly 10% compared to baseline models. © 2022, CC BY-NC-ND.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：