检索结果-内蒙古大学图书馆

Bilingual Lexicon Induction for Low-Resource Languages using Graph Matching via Optimal Transport

学校读者我要写书评

暂无评论

Bilingual Lexicon Induction for Low-Resource Languages using...

2022 Conference on Empirical Methods in Natural Language Processing, EMNLP 2022

作者： Marchisio, Kelly Saad-Eldin, Ali Duh, Kevin Priebe, Carey Koehn, Philipp Department of Computer Science Johns Hopkins University United States Department of Applied Mathematics and Statistics Johns Hopkins University United States Department of Biomedical Engineering Johns Hopkins University United States Human Language Technology Center of Excellence Johns Hopkins University United States

Bilingual lexicons form a critical component of various natural language processing applications, including unsupervised and semisupervised machine translation and crosslingual information retrieval. We improve bilingual lexicon induction performance across 40 language pairs with a graph-matching method based on optimal transport. The method is especially strong with low amounts of supervision. © 2022 Association for Computational Linguistics.

关键词： Natural language processing systems

Generative Adversarial Networks for Imputing Sparse Learning Performance

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zhang, Liang Yeasin, Mohammed Lin, Jionghao Havugimana, Felix Hu, Xiangen Institute for Intelligent Systems University of Memphis MemphisTN38152 United States Department of Electrical and Computer Engineering University of Memphis MemphisTN38152 United States Human-Computer Interaction Institute Carnegie Mellon University PittsburghPA15213 United States Centre for Learning Analytics Monash University MelbourneVIC3800 Australia Department of Applied Social Sciences Hong Kong Polytechnic University Hong Kong

Learning performance data, such as correct or incorrect responses to questions in Intelligent Tutoring Systems (ITSs) is crucial for tracking and assessing the learners’ progress and mastery of knowledge. However, the issue of data sparsity, characterized by unexplored questions and missing attempts, hampers accurate assessment and the provision of tailored, personalized instruction within ITSs. This paper proposes using the Generative Adversarial Imputation Networks (GAIN) framework to impute sparse learning performance data, reconstructed into a three-dimensional (3D) tensor representation across the dimensions of learners, questions and attempts. Our customized GAIN-based method computational process imputes sparse data in a 3D tensor space, significantly enhanced by convolutional neural networks for its input and output layers. This adaptation also includes the use of a least squares loss function for optimization and aligns the shapes of the input and output with the dimensions of the questions-attempts matrices along the learners’ dimension. Through extensive experiments on six datasets from various ITSs, including AutoTutor, ASSISTments and MATHia, we demonstrate that the GAIN approach generally outperforms existing methods such as tensor factorization and other generative adversarial network (GAN) based approaches in terms of imputation accuracy. This finding enhances comprehensive learning data modeling and analytics in AI-based education. © 2024, CC BY.

关键词： Generative adversarial networks

Big Data-Driven Strategies for Growth in 3PL with Advanced IT Solutions 1

学校读者我要写书评

暂无评论

Big Data-Driven Strategies for Growth in 3PL with Advanced I...

1st International Conference on Advanced Computing and Emerging Technologies, ACET 2024

作者： Yogi, Kottala Sri Suganthi, N. Dankan Gowda, V. Hanumantharao, Attru Srinivas, Ved Kavitha, R. Hyderabad Symbiosis International University Symbiosis Institute of Business Management Department of Operations Pune India Dayananda Sagar College of Engineering Department of Electrical and Electronics Engineering Karnataka Bangalore India Bms Institute of Technology and Management Department of Electronics and Communication Engineering Karnataka Bangalore India Aditya College of Engineering and Technology Department of Computer Science & Engineering Andhra Pradesh Surampalem India Thiagarajar School of Management Department of Human Resources Tamil Nadu Madurai India Velammal College of Engineering and Technology Department of Information Technology Madurai India

ISBN: (纸本)9798350367720

The focus of this paper is to examine how different big data initiatives support the development and operations of third party logistics providers. Focusing on the use of advanced technologies like IoT, AI, and ML in the context of 3PL, the research has the objective to show how these innovations enhance efficiency, operating cost, and customer satisfaction. Information gathered from live monitoring, use of algorithms or predictive techniques, and integration of logistics optimization techniques are used to highlight performance enhancement. According to the proposed model, the processes related to the utilisation of big data and IT solutions are divided into several categories: big data acquisition, cloud computing, and subsequent constant evaluation. This study reveals insights that 3PL providers can create a sustainable competitive advantage to the clientele firms by adopting accurate demand forecasting, improvement of routes, and optimization of supply chain visibility. Thus, the findings of this study add to the current literature and practice by enumerating effective big data technologies to apply in logistics management and showcasing how these advancements can catalyze sustainable development. © 2024 IEEE.

关键词： 3PL Artificial Intelligence Big Data Data Analytics Internet of Things IT Solutions Supply Chain Management Third-Party Logistics

Impact of Machine Learning on Applying the Best Worst Method for Social Sustainability in Manufacturing Supply Chains 7

学校读者我要写书评

暂无评论

Impact of Machine Learning on Applying the Best Worst Method...

7th International Conference on Circuit Power and Computing Technologies, ICCPCT 2024

作者： Yogi, Kottala Sri Dankan Gowda, V. Kumar Sahu, Atul Jagtap, Madan Mohanrao Srinivas, Ved Abdullah, Ibrahim International University Symbiosis Institute of Business Management Hyderabad Symbiosis Department of Operations Pune India BMS Institute of Technology and Management Department of Electronics and Communication Engineering Karnataka Bangalore India School of Studies in Engineering & Technology Department of Industrial and Production Engineering Koni Chattisgarh Bilaspur India Symbiosis Institute of Operations Management Department of Operations Management Nashik Campus Maharashtra Pune India Thiagarajar School of Management Department of Human Resources Tamil Nadu Madurai India Al-Turath University College Department of Computer Science Baghdad Iraq

ISBN: (纸本)9798350372816

This research is carried out merging ML with BWM to intensify decision making in manufacturing supply chain using social sustainability under light. As companies become more focused on environmental sustainability, the capability of machine learning to effectively assess and prioritize sustainability criteria should be highly valued. Through a team of domain experts, the research clearly identifies the most vital social sustainability factors and, afterwards, deploys ML algorithms to analyze and optimize the provided decision-making framework by BWM. The selected approach gives possibility to dynamic weighting of the criteria which is based on real-time data. This way we are able to come up with the more precise and effective sustainability strategies. The results show that the hybrid ML-BWM model is not only an effective tool to enhance the validity and accuracy of evaluations but also one that provides the prudent directions that help the manufacturing sector shape its choice and responses. This paper analyzes the possible outcome of the integration of AI and traditional approach to decision-making process on social aspects of sustainability within supply chain. © 2024 IEEE.

关键词： Smart manufacturing

Green IoT Edge Computing Towards Sustainable and Distributed Data Processing

学校读者我要写书评

暂无评论

Green IoT Edge Computing Towards Sustainable and Distributed...

IEEE UP Section Conference on Electrical computer and Electronics (UPCON)

作者： K.R. Ananth Hari Shankar Punna Krishnamoorthy Selvaraj Rajagopal K Vaishali Mahajan Sakthivel S. Department of Computer Science (Artificial Intelligence and Data Science) Nandha Arts and Science College Erode Tamilnadu Department Of CSE CVR College Of Engineering (An Autonomous Institution) Telangana Department of ERP Keller Marketing Symbiosis Centre for Management & Human Resource Development (SCMHRD) Pune Maharashtra Marketing Symbiosis Centre for Management & Human Resource Development (SCMHRD) Pune Maharashtra Department of Computer Applications K.S.Rangasamy College of Arts and Science Tiruchengode Tamilnadu

Data creation in the network's periphery has increased exponentially as IoT devices have become more commonplace. To address the challenges this large data deluge poses, sustainable and efficient data processing technologies are required. The term “Green IoT Edge Computing” refers to a new strategy that uses distributed data processing and renewable energy to improve the efficiency and longevity of Internet of Things (IoT) applications. In the Internet of Things (IoT), edge computing is used to process data locally, at the network's periphery, rather than forwarding it to a centralized data center. This method improves real-time decision-making, minimizes network latency, and helps preserve bandwidth. However, in locations with limited access to power or other resources, the high power consumption of edge computing devices continues to be a major issue. Green IoT edge computing addresses this challenge by combining renewable energy sources like solar panels and wind turbines with traditional edge computing facilities. This guarantees low- impact and long-term power for edge devices, cutting down on their environmental impact and operational expenses. This strategy harmonizes IoT operations with environmental sustainability targets by utilizing renewable energy sources. The ability to handle data in a distributed manner is also crucial to green IoT edge computing. Rather than a centralized server, a distributed group of edge devices processes data. This improves energy efficiency as well as fault tolerance and dependability. Edge devices working together to properly spread processing activities can reduce the demand for powerful, energy-hungry servers. Green IoT Edge Computing's primary invention is the “Renewable Edge Optimizer.” This clever piece of code automatically adjusts data processing tasks in accordance with the availability of renewable power. When clean energy is plentiful, the optimizer may shift more data processing to distributed nodes, relieving press

关键词：

DRDM: A Disentangled Representations Diffusion Model for Synthesizing Realistic Person Images

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Huang, Enbo Zhang, Yuan Huang, Faliang Zhang, Guangyu Liu, Yang Guangxi Key Lab of Human-machine Interaction and Intelligent Decision Nanning Normal University Nanning China Zhidayuan AI Lab. Nanning China College of Mathematics and Informatics South China Agricultural University Guangzhou China School of Computer Science and Engineering Sun Yat-Sen University Guangzhou China

Person image synthesis with controllable body poses and appearances is an essential task owing to the practical needs in the context of virtual try-on, image editing and video production. However, existing methods face significant challenges with details missing, limbs distortion and the garment style deviation. To address these issues, we propose a Disentangled Representations Diffusion Model (DRDM) to generate photo-realistic images from source portraits in specific desired poses and appearances. First, a pose encoder is responsible for encoding pose features into a high-dimensional space to guide the generation of person images. Second, a body-part subspace decoupling block (BSDB) disentangles features from the different body parts of a source figure and feeds them to the various layers of the noise prediction block, thereby supplying the network with rich disentangled features for generating a realistic target image. Moreover, during inference, we develop a parsing map-based disentangled classifier-free guided sampling method, which amplifies the conditional signals of texture and pose. Extensive experimental results on the Deepfashion dataset demonstrate the effectiveness of our approach in achieving pose transfer and appearance control. Copyright © 2024, The Authors. All rights reserved.

关键词： Signal encoding

PROM: A Phrase-level Copying Mechanism with Pre-training for Abstractive Summarization

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Ma, Xinbei Gong, Yeyun He, Pengcheng Zhao, Hai Duan, Nan Department of Computer Science and Engineering Shanghai Jiao Tong University China Key Laboratory of Shanghai Education Commission for Intelligent Interaction and Cognitive Engineering Shanghai Jiao Tong University China Microsoft Research Asia China

Based on the remarkable achievements of pretrained language models in abstractive summarization, the copying mechanism has proved helpful by improving the factuality, stability, and overall performance. This work proposes PROM, a new PhRase-level cOpying Mechanism that enhances attention on n-grams, which can be applied to zero-shot summarization with pre-training. PROM adds an indicator layer to explicitly pick up tokens in n-gram that can be copied from the source, and calculates an auxiliary loss for the copying prediction. Empirical studies show that PROM makes significant improvements in fine-tuning on benchmarks. In the zero-shot setting, PROM is utilized in the self-supervised pre-training on raw corpora and provides new general baselines on a wide range of summarization datasets. Further analysis shows that PROM performs more reasonable copying and contributes to faithfulness. Our code is publicly available at https://***/xbmxb/PROM. Copyright © 2023, The Authors. All rights reserved.

关键词： Zero-shot learning

ROBUST KNOWLEDGE DISTILLATION FROM RNN-T MODELS WITH NOISY TRAINING LABELS USING FULL-SUM LOSS

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Zeineldeen, Mohammad Audhkhasi, Kartik Baskar, Murali Karthick Ramabhadran, Bhuvana Human Language Technology and Pattern Recognition Computer Science Department Rwth Aachen University Aachen52074 Germany Google Llc New York United States

This work studies knowledge distillation (KD) and addresses its constraints for recurrent neural network transducer (RNNT) models. In hard distillation, a teacher model transcribes large amounts of unlabelled speech to train a student model. Soft distillation is another popular KD method that distills the output logits of the teacher model. Due to the nature of RNN-T alignments, applying soft distillation between RNNT architectures having different posterior distributions is challenging. In addition, bad teachers having high word-error-rate (WER) reduce the efficacy of KD. We investigate how to effectively distill knowledge from variable quality ASR teachers, which has not been studied before to the best of our knowledge. We show that a sequence-level KD, full-sum distillation, outperforms other distillation methods for RNN-T models, especially for bad teachers. We also propose a variant of full-sum distillation that distills the sequence discriminative knowledge of the teacher leading to further improvement in WER. We conduct experiments on public datasets namely SpeechStew and LibriSpeech, and on in-house production data. Copyright © 2023, The Authors. All rights reserved.

关键词： Recurrent neural networks

Research on Autonomous Robots Navigation based on Reinforcement Learning

学校读者我要写书评

暂无评论

Research on Autonomous Robots Navigation based on Reinforcem...

Robotics, Artificial Intelligence and Intelligent Control (RAIIC), International Conference on

作者： Zixiang Wang Hao Yan Zhuoyue Wang Zhengjia Xu Zhizhong Wu Yining Wang College of Engineering and Computer Science Syracuse University Syracuse NY Human Factors in Information Design Bentley University Waltham MA SATM Cranfield University Cranfield UK Department of Electrical Engineering and Computer Sciences University of CaliforniaBerkeley Berkeley CA Independent Researcher Mountain View CA

ISBN: (数字)9798350352627

ISBN: (纸本)9798350352634

Reinforcement learning continuously optimizes decision-making based on real-time feedback reward signals through continuous interaction with the environment, demonstrating strong adaptive and self-learning capabilities. In recent years, it has become one of the key methods to achieve autonomous navigation of robots. In this work, an autonomous robot navigation method based on reinforcement learning is introduced. We use the Deep Q Network (DQN) and Proximal Policy Optimization (PPO) models to optimize the path planning and decision-making process through the continuous interaction between the robot and the environment, and the reward signals with real-time feedback. By combining the Q-value function with the deep neural network, deep Q network can handle high-dimensional state space, so as to realize path planning in complex environments. Proximal policy optimization is a strategy gradient-based method, which enables robots to explore and utilize environmental information more efficiently by optimizing policy functions. These methods not only improve the robot’s navigation ability in the unknown environment, but also enhance its adaptive and self-learning capabilities. Through multiple training and simulation experiments, we have verified the effectiveness and robustness of these models in various complex scenarios.

关键词： Training Adaptation models Navigation Decision making Reinforcement learning Real-time systems Path planning