检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Wang, Xiaolong Zhang, Yuanchi Wang, Ziyue Xu, Yuzhuang Luo, Fuwen Wang, Yile Li, Peng Liu, Yang Dept. of Comp. Sci. & Tech. Institute for AI Tsinghua University Beijing China Tsinghua University Beijing China Gansu China Harbin Institute of Technology Harbin China College of Computer Science and Software Engineering Shenzhen University Shenzhen China

Large language models (LLMs) have revolutionized the field of natural language processing, enabling remarkable progress in various tasks. Different from objective tasks such as commonsense reasoning and arithmetic question-answering, the performance of LLMs on subjective tasks is still limited, where the perspective on the specific problem plays crucial roles for better interpreting the context and giving proper response. For example, in certain scenarios, LLMs may perform better when answering from an expert role perspective, potentially eliciting their relevant domain knowledge. In contrast, in some scenarios, LLMs may provide more accurate responses when answering from a third-person standpoint, enabling a more comprehensive understanding of the problem and potentially mitigating inherent biases. In this paper, we propose Reasoning through Perspective Transition (RPT), a method based on in-context learning that enables LLMs to dynamically select among direct, role, and third-person perspectives for the best way to solve corresponding subjective problem. Through extensive experiments on totally 12 subjective tasks by using both closed-source and open-source LLMs including GPT-4, GPT-3.5, Llama-3, and Qwen-2, our method outperforms widely used single fixed perspective based methods such as chain-of-thought prompting and expert prompting, highlights the intricate ways that LLMs can adapt their perspectives to provide nuanced and contextually appropriate responses for different problems. © 2025, CC BY.

关键词： Open systems

来源：评论

学校读者我要写书评

暂无评论

Deep Learning-Enhanced Medical System for Disease Prediction, Medication Recommendation, and Intelligent Question-Answering Based on Patient Symptomatology

Deep Learning-Enhanced Medical System for Disease Prediction...

引用

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

作者： Kabiru Uba Kiru Surya Kant Pal Salim Saad Ahmad Ismail Muhammad Kabir Department of Mathematics School of Basic Sciences and Research Sharda University Greater Noida Uttar Pradesh India Department of Computer Sci. School of Engineering and Technology Sharda University Greater Noida Uttar Pradesh India

ISBN: (数字)9798350378726

ISBN: (纸本)9798350378733

Deep Learning (DL) is branch of machine learning that emulates the architecture and functioning of the human brain through neural networks to process and analyzes large amounts of data and extract meaningful patterns and insight. This study presents an innovative medical system that leverages deep learning and natural language processing (NLP) techniques to predict diseases based on symptomatology and provide personalized health suggestions, which involves an indepth design of a deep learning-driven medical assistant platform, which will feature the integration of state-of-the-art deep learning models with multiple NLP techniques and models for predicting the disease that a patient may have just by describing the symptoms, prescribing medication, and answering medical-related questions. This project provides a long lasting solutions to the various healthcare issues, like; assisting medical professionals in decision-making quickly, likewise assists patient with answers to medical queries, also assists patient by providing instantaneous, accurate, and personalized health suggestions with pending approval from the doctors and professionals. Bidirectional Encoder Representations from Transformers (BERT) like models will be adopted to answer questions and classify diseases together Long Short-Term Memory (LSTM) for classification task, with BERT also for the remaining tasks. Health access is further facilitated while reducing chances of misdiagnosis that lead to bad outcomes for patients because of intelligent automation. The use of machine learning models allows adaptation to new data, and therefore the system is robust and scalable. The study utilizes TF-IDF vectorization with Hugging Face Transformers, PyTorch, sci.it-learn, Pandas, NLTK libraries, hugging face trainer class and also accuracy, precision, recall f1-score and exact match metrics. The models attained high scores in various metrics which make all the tasks efficient to use by both medical and non-medical

关键词： Deep learning Accuracy Medical services Bidirectional control Predictive models Transformers Natural language processing Encoding Medical diagnostic imaging Diseases

来源：评论

学校读者我要写书评

暂无评论

Boosting the Prediction of Brain Tumor Using Two Stage BiGait Architecture

Boosting the Prediction of Brain Tumor Using Two Stage BiGai...

引用

Proceedings of the Digital Image Computing: Technqiues and Applications (DICTA)

作者： Saif-ur-Rehman Khan Zia Khan Md Zakir Hossain Nicanor Mayumu Farhana Yasmin Younas Aziz School of Computer Science and Engineering Central South University Changsha China School of Elec Eng Comp and Math Sci (EECMS) Curtin University Perth Australia Department of Computer and Software Nanjing University of information Science and Technology Nanjing

ISBN: (数字)9798350379037

ISBN: (纸本)9798350379044

Deep learning (DL) techniques hold immense promise for revolutionizing medical diagnostics, including brain tumor detection. Detecting malignancies in the brain is fraught with challenges that carry critical implications for accurate and timely diagnosis. Early detection is paramount, and our paper presents a novel approach to address this issue. We introduce the BiGait ensemble transfer learning (TL) architecture tailored for brain tumor classification tasks. Our proposed methodology is a two-stage process that harnesses the power of TL. In the initial phase, we employ two established models, ResNet50 and EfficientNetB1, and fine-tune them using the Kaggle multiclass dataset. This includes the incorporation of supplementary feature layers. We also leverage a stack ensemble solution to combine these feature layers, enhancing the model's overall predictive capabilities. Our approach is designed to shift the input distribution positively, effectively mitigating overfitting and reducing hardware overhead. To validate the effectiveness of our method, we conducted extensive experiments and evaluation. The dataset utilized in this work, publicly available on Kaggle contains 3,264 brain MRIs categorized into four groups: normal (500 images), meningioma (937 images), glioma (926 images), and pituitary tumor (901 images). The experimental results demonstrate that our model surpasses traditional single-model approaches, achieving an impressive accuracy rate of 96% and exhibiting superior generalization capabilities. This advancement in brain tumor classification holds great promise for early diagnosis and improved patient outcomes.

关键词： Accuracy Transfer learning Brain tumors computer architecture Predictive models Brain modeling Reliability Medical diagnosis Residual neural networks Overfitting

来源：评论

学校读者我要写书评

暂无评论

Discrete Shuffled Frog Leading Algorithm for Multiple-product Human-robot Collaborative Disassembly Line Balancing Problem

Discrete Shuffled Frog Leading Algorithm for Multiple-produc...

引用

2022 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2022

作者： Fan, ChenYang Wang, JiaCun Guo, XiWang Qin, ShuJin Zhou, MengChu Qi, Liang Engineering of Liaoning Petrochemical University School of Information and Control Fushun China Engineering of Monmouth University Dept. Computer Science and Software West Long Branch United States Liaoning Petrochemical University Fushun China Monmouth University West Long Branch United States College of Economics and Management of Shangqiu Normal University Shangqiu China Dept. Electrical and Computer Engineering of New Jersey Institute of Technology Newark United States Tech of Shandong University of Science and Technology Dept. Computer Intelligent Sci. & Qingdao China

ISBN: (数字)9781665452588

ISBN: (纸本)9781665452588

With the rapid development of recycling and remanufacturing technologies, disassembly line balancing problems (DLBP) have drawn great attention. Considering the limitation of disassembly by humans or robots alone, this paper focuses on human-robot collaborative disassembly lines. Specifically, this work proposes a multi-product human-robot collaborative disassembly line balancing model to tackle the inflexibility of single product disassembly and inconsistency in recovery values of different product components. Its objective is to maximize disassembly profit. As a commercially available solver, IBM's CPLEX is used to obtain the exact solution of DLBP and verify the correctness of the proposed mathematical model. A discrete shuffled frog leading algorithm is newly designed to solve the sizable problems. Experimental results show that the proposed algorithm has a fast convergence rate and can find solutions consistent with those with CPLEX but requires much less time than the latter, thus advancing the field of disassembly automation. © 2022 IEEE.

关键词： Robots

来源：评论

学校读者我要写书评

暂无评论

Work in Progress: VR-based Robotics Training for AEC Industry

Work in Progress: VR-based Robotics Training for AEC Industr...

引用

IEEE World engineering Education Conference (EDUNINE)

作者： Biayna Bogosian Shahin Vassigh Bhavleen Kaur Narula Seth Corrigan Giancarlo Perez Mohammadreza Akbari Lor Bhanu Vodinepally Tisa Islam Erana Mark Alan Finlayson Shu-Ching Chen The Design School Arizona State University Tempe AZ USA Department of Architecture Florida International University Miami FL USA School for Inf. and Computer Sci. University of California Irvine CA USA School of Science and Engineering University of Missouri Kansas City MO USA School of Computer and Inf. Sci. Florida International University Miami FL USA

ISBN: (数字)9798331542788

ISBN: (纸本)9798331542795

This paper introduces the Intelligent Learning Platform for Robotics Operations (IL-PRO), a Virtual Reality (VR) system designed to enhance robotics training in the Architecture, engineering, and Construction (AEC) industry. IL-PRO addresses the growing need for effective training methods as the AEC sector adopts robotic automation. The system integrates VR technology with game-assisted learning, combining online multimedia lessons for theory with immersive VR tasks for practical skills. Developed iteratively using Design-Based Research principles, IL-PRO incorporates realistic robot simulations and progressive task complexity. The VR environment, built in Unity, aims to enhance engagement, motor coordination, and spatial awareness in robotics training. While future goals include AI-driven personalized instruction, this work-in-progress focuses on VR curriculum development and implementation. The paper concludes by discussing future directions, including curriculum expansion and cross-institutional adoption, to establish new benchmarks in innovative robotics education for the AEC industry.

关键词： Training Industries Solid modeling Service robots Robot kinematics Multimedia systems Curriculum development Motor coordination engineering education Immersive learning

来源：评论

学校读者我要写书评

暂无评论

Investigating Pre- and Post-Seismic TEC Anomalies Related to Moderate Earthquakes in Türkiye

Investigating Pre- and Post-Seismic TEC Anomalies Related to...

引用

2023 4th International Informatics and software engineering Conference (IISEC)

作者： Ahmad Muhammad Fatih Külahçı Asaf Varol Salim Jibrin Danbatta Department of Physics and Material Sci. College of Arts and Sciences Qatar University Doha Qatar Physics Department Nuclear Physics Division Firat University Elazığ Türkiye Dept. of Engineering Mgmt. & Tech. College of Engineering and Comp. Sci. The University of Tennessee at Chattanooga Chattanooga TN United States Software Engineering Department Uskudar University Istanbul Türkiye

GPS Total Electron Content (GPS TEC), variations are analyzed over Türkiye for the periods from the 17th of August 2008 to the 21st of September 2008. This analysis was conducted in quite solar and geomagnetic conditions. The standard deviation and seasonal decomposition methods are adapted to study the TEC time series anomalies. For the two major earthquake events with magnitudes 4.7 and 5.1, spatial TEC disturbances are visualized by plotting contoured TEC maps. Shallow earthquakes are considered with a magnitude range. Anomaly detection results suggest the significance of moderate earthquakes in seismic-ionospheric studies.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Design and Enhancement of a CNN Model to Augment the Face Recognition Accuracy 3

Design and Enhancement of a CNN Model to Augment the Face Re...

引用

3rd International Informatics and software engineering Conference, IISEC 2022

作者： Kak, Shakir Fattah Mustafa, Firas Mahmood Varol, Asaf Duhok Polytechnic University Akre Technical Institutes Information Technology Department Duhok Iraq Duhok Polytechnic University Technical College of Engineering Chemical Engineering Dept. Iraq College of Engineering Nawroz University Department of Computer and Communication Duhok Iraq The University of Tennessee Dept. of Engineering Mgmt. & Tech. College of Engineering and Comp. Sci. ChattanoogaTN United States

ISBN: (数字)9781665459952

ISBN: (纸本)9781665459952

In the last decade, smart home security applications have relied more on human biometrics in their functions, due to the reliability and the high-precision results these technologies provide. Face recognition is one of the popular biometrics in the field of image processing technologies. Human face recognition processing is a complicated operation that involves different factors and circumstances such as the illumination degree and the position of the face that affects the final recognition rate. In this research, the Convolution Neural Network (CNN) architecture is used in the extracting phase of significant features of the face shape, and the SoftMax classification layer was used to identify faces in the fully connected CNN layer. This paper provides an update of CNN architecture by applying a three-batch normalization layer to the CNN design. By applying this modification, the system network speed increased with a better recognition rate. The recognition rate also increased by applying two DWT levels with a bio5.5 filter to the training group of the database images and the tested image before applying the PCA dimensional reduction algorithm instead of using the PCA algorithm alone. The obtained face recognition rates have been improved to 99.75% by applying the proposed CNN scheme. While applying the proposed hybrid approach (using the PCA next to applying DWT-2 levels with bior5.5 filter) has registered a 99.25% recognition rate compared to a 96.75% recognition rate when obtained by applying the PCA method alone. The research has adopted using a set of 360 training images and 40 test images set of the standard ORL Database in its work. © 2022 IEEE.

关键词： Biometrics

来源：评论

学校读者我要写书评

暂无评论

Manipulation and Measurement of Knowledge Representations of Language Models 2

Manipulation and Measurement of Knowledge Representations of...

引用

2nd International Conference on Self Sustainable Artificial Intelligence Systems, ICSSAS 2024

作者： Garg, Raghav Manu, Mehul Charu Chauhan, Neha Naval, Naveen Tharini, C. Tula's Institute Dept of Computer Sci & Engineering Dehradun India Graphic Era Hill University Dept of Allied Sciences Bhimtal India Dev Bhoomi Uttarakhand University Deptt. of Electrical Engg. Dehradun India Graphic Era Deemed to be University Deptt. of Humanities & Social Sci. Dehradun India Vel Tech Rangarajan Dr Sagunthala R&d Institute of Science and Technology Department of English. Chennai India

ISBN: (纸本)9798350368413

Modern neural language models (LMs) are often pre-trained with a self-supervised objective: are given with texts that have piece(s) withheld, and asked to generate the withheld portions of the text. Large amounts of factual knowledge can be stored by language models but it is yet unclear if these models can apply this knowledge to logical reasoning. Tasks that come after may benefit from the factual knowledge that was gathered during pretraining and saved in Language Models' (LMs') parameters. The study introduces BART, a technique for learning to translate natural language claims into fact encodings stored in the internal representation system of an LM. BART leverages architecture of denoising autoencoder to refine and manipulate language by pre-training and fine-tuning for effective management of text through utilization of semantic and linguistic knowledge. BART encodings can be employed as knowledge editors to adjust subsequent generation to be compatible with freshly learned facts when applied to LM concealed representations. Because BART encodings reveal which attributes LMs already attribute to the indicated entities, it can also be used as probes when compared to LM representation. In certain situations, this allows for the prediction of when LMs will produce outputs that are at odds with input text or background knowledge. BART provides a link between the work on probing, prompting, and LM editing, as well as a path towards universal tools for fine-grained knowledge control and inspection in LMs. The research gap identified is effective application of knowledge in language models to enhance logical reasoning and maintaining coherence using BART through effective application of acquired knowledge. © 2024 IEEE.

关键词： Knowledge representation

来源：评论

学校读者我要写书评

暂无评论

Machine Learning Based Automated Approach to Detect Brain Disease Anomalies

Machine Learning Based Automated Approach to Detect Brain Di...

引用

2021 International Conference on Computational Intelligence and Computing Applications, ICCICA 2021

作者： Dubey, Shatrughan Pandey, Yogadhar Technocrats Inst. of Technol. and Sci. Proudyogiki Vishwavidyalaya Rajiv Gandhi Proudyogiki Vishwavidyalaya Department of Computer Science i Engineering Bhopal India Technocrats Institute of Technology and Science Rajiv Gandhi Department of Computer Science and Engineering Bhopal India

ISBN: (纸本)9781665420402

This paper proposed a new model which isi based oni the classification methods such asi support vector machine neurali network andi optimization methods which isi bi-logically inspired method for the improving the classifier results in the terms ofisome performance parameters such as accuracy, precision, recall etc., here we measure the all performance parameters for the various dataset such as heart patients, liver patients andi cancer patients and improve the rate of classification or resultsi with compare than other existing techniques. The alli patient's dataset whichi is taken fromitheiuci machine learning repository whichi providei the authentic dataset for the research work and thei simulation software isimatlab. Ini thisi paper our experimental results shows thati theibetter detectioniratei of classification for performance parameters thani other existingi techniques. © 2021 IEEE.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

Recommendation System in Industrial data analysis using Machine Learning Techniques 3

Recommendation System in Industrial data analysis using Mach...

引用

3rd IEEE North Karnataka Subsection Flagship International Conference, NKCon 2024

作者： Joshi, Tripuresh Joshi, Bhavesh Syed, Karimulla Vijayarani, S. Kumar, Sandeep Naval, Naveen Tula's Institute Deptt. of Electronics & Comm. Engg. UK Dehradun India Graphic Era Hill University Scool of Management India University of Technology and Applied Sciences Shinas College of Engineering and Technology Engineering Department Oman Srm Institute of Science and Technology Department of Mba Chennai India Tula's Institute Deptt. of Computer Science & Engg. U.K. Dehradun India Graphic Era Deemed to Be University Deptt. of Humanities & Social Sci. Dehradun India

ISBN: (纸本)9798350364569

The interactions of users are seamlessly integrated into the product life cycle in the Industry structure of the future. Recommender systems function as a means of filtering information by providing users with recommendations according to their interests in items, music, friends, and topics. Numerous research endeavours have aimed to enhance recommendation accuracy through various techniques like matrix factorization, trust-based approaches, hybrids, and both machine and deep learning methods. However, few studies have explored the utilization of textual opinions for recommendations. Existing research primarily relies on numerical ratings, which may not accurately represent user behavior. This study delves into analyzing the sentiments expressed in textual opinions to gain a deeper understanding of user behavior. Furthermore, textual opinions are processed through the use of methods from Natural Language Processing such as stop-word removal, lemmatization, stemming, and Part-of-Speech (POS) tagging. The aim of a recommendation system is to anticipate user preferences and understand their thought patterns. By considering user preferences and demands, the system delivers tailored information. A deeper examination of the data is necessary to enhance recommendation quality. © 2024 IEEE.

关键词： Matrix factorization

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：