检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Kim, Seongho Moon, Jihyun Oh, Juntaek Choi, Insu Yang, Joon-Sung Department of Electrical and Electronic Engineering Yonsei University Seoul Korea Republic of Department of Systems Semiconductor Engineering Yonsei University Seoul Korea Republic of BK21 Graduate Program in Intelligent Semiconductor Technology Yonsei University Seoul Korea Republic of

Large Language Models (LLMs), emerging from advancements in Natural Language Processing (NLP) tasks, allow chatbots to provide more sophisticated and human-like text generation by leveraging their vast model sizes, often exceeding billions of parameters, to provide deep knowledge across various domains. Hence, they are becoming more integrated into our daily lives, serving as personal assistants or even as experts across various domains. Initial language models relied on rule-based systems and early neural networks like Recurrent Neural Networks (RNNs). However, it remained an issue to handle long-term dependency and to understand context over extended conversation. The advent of the Attention mechanism and Transformer architecture enables contextually natural text generation and compresses the burden of processing entire source information into singular vectors. Based on these two main ideas, model sizes gradually increases to accommodate more precise and comprehensive information, leading to the current state-of-the-art LLMs being very large, with parameters around 70 billion. As the model sizes are growing, the demand for substantial storage and computational capacity increases. This leads to the development of high-bandwidth memory and accelerators, as well as a variety of model architectures designed to meet these requirements. We note that LLM architectures have increasingly converged. This paper analyzes how these converged architectures perform in terms of layer configurations, operational mechanisms, and model sizes, considering various hyperparameter settings. In this paper, we conduct a concise survey of the history of LLMs by tracing the evolution of their operational improvements. Furthermore, we summarize the performance trends of LLMs under various hyperparameter settings using the RTX 6000, which features the state-of-the-art Ada Lovelace architecture. We conclude that even the same model can exhibit different behaviors depending on the hyperparamete

关键词： Recurrent neural networks

来源：评论

学校读者我要写书评

暂无评论

Improvement of field phenotyping from synchronized multi-camera image collection based on multiple UAVs collaborative operation systems

Improvement of field phenotyping from synchronized multi-cam...

引用

2022 ASABE Annual International Meeting

作者： Lee, Hyeon-Seung Thomasson, J. Alex Han, Xiongzhe Department of Biosystems Engineering College of Agriculture and Life Sciences Kangwon National University Gangwon Chuncheon24341 Korea Republic of Interdisciplinary Program in Smart Agriculture College of Agriculture and Life Sciences Kangwon National University Gangwon Chuncheon24341 Korea Republic of Department of Agricultural and Biological Engineering Mississippi State University StarkvilleMS39762 United States

The technology using multiple unmanned aerial vehicles has been intensively applied in the national defense and the festival (event) but has not been carried out to improve the agricultural field. To successfully implement precision agriculture for efficient crop production, three-dimensional (3D) reconstruction of the plant canopy is essential to measure geometry such as height, width, volume, and leaf coverage. However, image data collection based on a single unmanned aerial vehicle is challenging to acquire an accurate crop shape depending on the angle of sunlight, and it is difficult to accurately measure leaves and stems in an environment in which crops are grown closely. To improve the 3D models to minimize errors caused by wind and lack of detailed information in overlapping regions, an image-based 3D plant reconstruction system based on multiple unmanned aerial vehicles is proposed to achieve two images simultaneously from different views of the plants during flight. An unresolvable offset may produce a stereo view of the environment if unsynchronized cameras capture nonstationary objects. Therefore, camera synchronization is required for the simultaneous acquisition of images from different perspectives to compensate for errors in plant-height measurement. This study will reconstruct 3D models of the crop with Structure-from-Motion based on a multiple-view stereo algorithm and metric construction and compare errors between traditional 3D models from single-camera image collection with errors from the synchronized multi-camera image collection. © 2022 ASABE. All Rights Reserved.

关键词： Unmanned aerial vehicles (UAV)

来源：评论

学校读者我要写书评

暂无评论

Stage-Wise Reward Shaping for Acrobatic Robots: A Constrained Multi-Objective Reinforcement Learning Approach

arXiv

引用

arXiv 2024年

作者： Kim, Dohyeong Kwon, Hyeokjin Kim, Junseok Lee, Gunmin Oh, Songhwai The Department of Electrical and Computer Engineering and ASRI Seoul National University Seoul08826 Korea Republic of The Interdisciplinary Program in Artificial Intelligence and ASRI Seoul National University Seoul08826 Korea Republic of

As the complexity of tasks addressed through reinforcement learning (RL) increases, the definition of reward functions also has become highly complicated. We introduce an RL method aimed at simplifying the reward-shaping process through intuitive strategies. Initially, instead of a single reward function composed of various terms, we define multiple reward and cost functions within a constrained multi-objective RL (CMORL) framework. For tasks involving sequential complex movements, we segment the task into distinct stages and define multiple rewards and costs for each stage. Finally, we introduce a practical CMORL algorithm that maximizes objectives based on these rewards while satisfying constraints defined by the costs. The proposed method has been successfully demonstrated across a variety of acrobatic tasks in both simulation and real-world environments. Additionally, it has been shown to successfully perform tasks compared to existing RL and constrained RL algorithms. Our code is available at https://***/rllab-snu/Stage-Wise-CMORL. Copyright © 2024, The Authors. All rights reserved.

关键词： Cost functions

来源：评论

学校读者我要写书评

暂无评论

Multi-Objective Grouped Grey Wolf Optimization of PID Controllers Applied to a Water Treatment Plant Model

Multi-Objective Grouped Grey Wolf Optimization of PID Contro...

引用

Congress on Evolutionary Computation

作者： Yan Lieven Souza Lúcio Viviana Cocco Mariani Leandro Dos Santos Coelho Electrical Engineering Graduate Program (PPGEE) Federal University of Parana (UFPR) Curitiba Parana Brazil Department of Electrical Engineering Mechanical Engineering Graduate Program (PPGEM) Pontifical Catholic University of Parana (PUCPR) and Federal University of Parana (UFPR) Curitiba Parana Brazil Department of Electrical Engineering Industrial and Systems Engineering Graduate Program (PPGEPS) Pontifical Catholic University of Parana (PUCPR) and Federal University of Parana (UFPR) Curitiba Parana Brazil

A comparison of proportional-integral-derivative (PID) controller tuning for a 2x2 Multiple-Input Multiple-Output (MIMO) water treatment plant model is presented. The PID tuning procedures are based on Multi-Objective Grey Wolf Optimizer (MOGWO) and Multi-Objective Grouped Grey Wolf Optimizer (MOGGWO) are performed under two different initial conditions. The tuning procedure is structured as a multi-objective optimization problem with four objective functions, the integral time-squared error is minimized in each output response, and the integral of the control actions is minimized in each input. MOGWO is a metaheuristic based on the hunting behavior of grey wolves, following the hierarchy and leadership procedures that these animals have, while the proposed MOGGWO is a new version proposed in this study with two different groups with different functionalities. The performance was analyzed by 50 runs based on statistical metrics. MOGGWO can find more local minima since its Pareto front is more dispersed than that found by MOGWO.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Determining Factors Affecting Customer Loyalty and Satisfaction in Online Food Delivery Service during the COVID - 19 Pandemic: A UTAUT2 Approach

Determining Factors Affecting Customer Loyalty and Satisfact...

引用

IEEE International Conference on Industrial engineering and engineering Management

作者： Krisna Chandra Susanto Yogi Tri Prasetyo Maela Madel L. Cahigas Reny Nadlifatin Satria Fadil Persada Irene Dyah Ayuwati Department of Industrial Engineering and Management Yuan Ze University Taoyuan Taiwan International Bachelor Program in Engineering Yuan Ze University Taoyuan Taiwan School of Industrial Engineering and Engineering Management Mapúa University Manila Philippines Department of Information Systems Institut Teknologi Sepuluh Nopember Surabaya Indonesia Entrepreneurship Department Bina Nusantara University Jakarta Indonesia University of Surabaya Surabaya Indonesia

ISBN: (数字)9798350386097

ISBN: (纸本)9798350386103

The Online food delivery services (OFDS) have been extensively utilized, especially in developing nations like Indonesia on new normal Covid-19 situation. The present research aimed to utilize the “Unified Theory of Acceptance and Use of Technology-2” (UTAUT2) in affecting the customer satisfaction and loyalty (CSL) in OFDS in Indonesia post-COVID-19 pandemic. The 253 data was taken to answer the 65 indicators. Structural equation modeling (SEM) indicates that hedonic motivation exerts the most significant influence on CSL, succeeded by price. This study unexpectedly found that usability factors, including performance expectation habit, did not influence customer happiness and CSL in OFDS. This research strongly provides a theoretical framework for OFDS practitioners, IT developers, and scholars. The present research can be modified and expanded to different nations to examine the determinants of customer CSL in OFDS.

关键词： COVID-19 Pandemics Numerical analysis engineering management Customer satisfaction Pricing Industrial engineering Mathematical models Usability

来源：评论

学校读者我要写书评

暂无评论

Customer Preferences for Online Games: A Conjoint Analysis Approach with Market Segmentation 22

Customer Preferences for Online Games: A Conjoint Analysis A...

引用

Proceedings of the 2022 6th International Conference on Software and e-Business

作者： Rogel Angelo Rebualos Yogi Tri Prasetyo Irene Dyah Ayuwati Satria Fadil Persada School of Industrial Engineering and Engineering Management Mapua University Philippines and School of Graduate Studies Mapua University Philippines School of Industrial Engineering and Engineering Management Mapua University Philippines and International Program in Engineering for Bachelor Yuan Ze University Taiwan Department of Information Systems Institut Teknologi Telkom Surabaya Indonesia Entrepreneurship Department BINUS Business School Undergraduate Program Indonesia

ISBN: (纸本)9798400700095

Online games display a unique market with a promising economy; hence many developer companies are exploring and have been utilizing this technology because of good possibility in recent years. This study aimed to determine the composition of online game attribute that is most preferred by customers using a Conjoint Analysis Approach. This study utilized different attributes such as graphics, player-to-player interaction, character creation freedom, game accessibility, monetization, storyline, and business model. An orthogonal design of conjoint analysis was utilized to assess customer preference for online games. A survey was proposed to a community of online game players and segmented among their gender, age, and income. The results showed that business model was the most considered attribute while accessibility and monetization were the least preferred. The findings of this study could be beneficial to both technology corporations developing products that fit their target customers and customers making most of their investment.

关键词： market segmentation preference conjoint online games

来源：评论

学校读者我要写书评

暂无评论

What Role Do Aggregators Play in Power System Security and Resilience?

What Role Do Aggregators Play in Power System Security and R...

引用

2023 Resilience Week, RWS 2023

作者： Matsuda-Dunn, Reiko Leddy, Laura Hotchkiss, Eliza Gautam, Mukesh Abdelmalak, Michael National Renewable Energy Laboratory GoldenCO80401 United States Colorado School of Mines Advanced Energy Systems PhD Program GoldenCO80401 United States University of Nevada at Reno Department of Electrical and Biomedical Engineering RenoNV89557 United States

ISBN: (纸本)9798350347470

Barriers to the participation of distributed energy resources (DERs) in wholesale electricity markets have limited the use of DERs for power system security and resilience. In September 2020, the Federal Energy Regulatory Commission (FERC) approved an order to reduce these barriers. FERC Order No. 2222 enables the participation of DER aggregators in wholesale electricity markets. DERs include renewable generation and technologies that support the integration of renewable generation by increasing grid flexibility and resilience. Requiring wholesale energy markets to allow DER aggregator participation provides a path for DERs to become competitive in these markets. As the contribution from aggregated DERs continues to increase, the aggregator's role in supporting grid security and resilience will become more critical. This paper reviews work that demonstrates how DER aggregators can provide resilience support through technical capabilities, operational strategies, and secure communication architectures. Socioeconomic influences and impacts of aggregators, including implications for social resilience, are presented. In surveying the current state-of-the-art across different but interconnected topics, we illustrate how aggregators can be power system participants that enhance grid security. There is no one-size-fits-all approach to enhancing resilience in a power grid that includes a growing cohort of DER aggregators, but there are many options for aggregators to contribute to a more resilient and secure power grid. © 2023 IEEE.

关键词： Power markets

来源：评论

学校读者我要写书评

暂无评论

Safe CoR: A Dual-Expert Approach to Integrating Imitation Learning and Safe Reinforcement Learning Using Constraint Rewards

arXiv

引用

arXiv 2024年

作者： Kwon, Hyeokjin Lee, Gunmin Lee, Junseo Oh, Songhwai Interdisciplinary Program in Artificial Intelligence ASRI Seoul National University Seoul08826 Korea Republic of Department of Electrical and Computer Engineering ASRI Seoul National University Seoul08826 Korea Republic of

In the realm of autonomous agents, ensuring safety and reliability in complex and dynamic environments remains a paramount challenge. Safe reinforcement learning addresses these concerns by introducing safety constraints, but still faces challenges in navigating intricate environments such as complex driving situations. To overcome these challenges, we present the safe constraint reward (Safe CoR) framework, a novel method that utilizes two types of expert demonstrations—reward expert demonstrations focusing on performance optimization and safe expert demonstrations prioritizing safety. By exploiting a constraint reward (CoR), our framework guides the agent to balance performance goals of reward sum with safety constraints. We test the proposed framework in diverse environments, including the safety gym, metadrive, and the real-world Jackal platform. Our proposed framework enhances the performance of algorithms by 39% and reduces constraint violations by 88% on the real-world Jackal platform, demonstrating the framework’s efficacy. Through this innovative approach, we expect significant advancements in real-world performance, leading to transformative effects in the realm of safe and reliable autonomous agents. Copyright © 2024, The Authors. All rights reserved.

关键词： Demonstrations

来源：评论

学校读者我要写书评

暂无评论

Use of Conjoint Analysis Filipino Consumer Preference on Attributes of Fixed Line Internet Service Providers 22

Use of Conjoint Analysis Filipino Consumer Preference on Att...

引用

Proceedings of the 2022 6th International Conference on Software and e-Business

作者： Justine Kyle Chan Nattakit Yuduang Yogi Tri Prasetyo Irene Dyah Ayuwati Satria Fadil Persada School of Industrial Engineering and Engineering Management Mapua University Philippines and School of Graduate Studies Mapua University Philippines School of Industrial Engineering and Engineering Management Mapua University Philippines and International Program in Engineering for Bachelor Yuan Ze University Taiwan Department of Information Systems Institut Teknologi Telkom Surabaya Indonesia Entrepreneurship Department BINUS Business School Undergraduate Program Indonesia

ISBN: (纸本)9798400700095

Internet has evolved from a basic communication medium into wide networks of communication networks where it has become integrated in people's daily lives. Not only that, because of the recent pandemic, the demand for internet has been rapidly increasing. With the integration of internet to our day to day interactions, choosing a provider that is inclined with consumers preference and usage should be the basis. This study aims to determine the Filipino consumer preference on choosing Fixed line internet service providers using Conjoint analysis. The method uses survey questionnaires that is disseminated across the regions of the Philippines. The data gathered from the survey are analyzed using SPSS. The result shows that out of all the attributes considered internet speed is the top priority of the consumers followed by pricing and brand. Overall, this study is beneficial in developing market strategies for companies in the telecommunications industry.

关键词： Internet Service Provider Internet ISP Conjoint Analysis Attributes

来源：评论

学校读者我要写书评

暂无评论

Large Language Model-Based Online Review Classification for Sub-Feature-Level Customer Opinion Analysis

SSRN

引用

SSRN 2024年

作者： Park, Seyoung Joung, Junegak Kim, Harrison Hyung Min Department of Industrial and Enterprise Systems Engineering University of Illinois Urbana-Champaign UrbanaIL61801 United States School of Interdisciplinary Industrial Studies Hanyang University Seoul04763 Korea Republic of

In recent years, many studies have analyzed online reviews to understand customer preferences and requirements for product features. However, most of them have focused on feature categories, whereas companies require design implications for sub-features. To bridge this gap between research and industry, this study developed a new methodology for sub-feature-level review analysis. First, text review sentences are embedded into vectors using a large language model. A sentence bidirectional encoder representation from transformer (SBERT) model is employed. Next, the method trains a neural network model that classifies reviews into sub-features. The input data are sentence vectors and the outputs are class labels indicating product subfeatures. To address the problem of highly imbalanced labels in review data, this study formulated a new loss function based on evaluation metrics. The developed methodology was tested using smartphone and headphone reviews collected online. The results showed that the new method achieved higher performance, i.e., F1 scores over 0.80, than a previous BERT-based classifier (F1 scores between 0.39 and 0.69). In addition, the new loss function provides a more balanced precision/recall for all the classes. The developed approach will help companies extract customer opinions at the product sub-feature level and has practical implications for early-stage product design. © 2024, The Authors. All rights reserved.

关键词： Product design

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：