检索结果-内蒙古大学图书馆

Advancements in Liver Tumor Detection:A Comprehensive Review of various Deep Learning Models

computer Modeling in engineering & sciences 2025年第1期142卷 91-122页

作者： shanmugasundaram Hariharan D.Anandan Murugaperumal Krishnamoorthy vinay Kukreja Nitin Goyal shih-Yu Chen Deptartment of Artificial Intelligence and Data Science Vardhaman College of EngineeringHyderabad501218India Deptartment of CSE V.S.B College of EngineeringKarur639111India Deptartment of Electrical and Electronics Engineering Vardhaman College of EngineeringHyderabad501218India Centre for Research Impact&Outcome Chitkara UniversityRajpura140401PunjabIndia Department of Computer Science and Engineering School of Engineering and TechnologyCentral University of HaryanaMahendergarh123031HaryanaIndia Department of Computer Science and Information Engineering National Yunlin University of Science and TechnologyYunlin64002Taiwan Intelligence Recognition Industry Service Research Center National Yunlin University of Science and TechnologyYunlin64002Taiwan

Liver cancer remains a leading cause of mortality worldwide,and precise diagnostic tools are essential for effective treatment *** Tumors(LTs)vary significantly in size,shape,and location,and can present with tissues of similar intensities,making automatically segmenting and classifying LTs from abdominal tomography images crucial and *** review examines recent advancements in Liver segmentation(Ls)and Tumor segmentation(Ts)algorithms,highlighting their strengths and limitations regarding precision,automation,and *** metrics are utilized to assess key detection algorithms and analytical methods,emphasizing their effectiveness and relevance in clinical *** review also addresses ongoing challenges in liver tumor segmentation and identification,such as managing high variability in patient data and ensuring robustness across different imaging *** suggests directions for future research,with insights into technological advancements that can enhance surgical planning and diagnostic accuracy by comparing popular *** paper contributes to a comprehensive understanding of current liver tumor detection techniques,provides a roadmap for future innovations,and improves diagnostic and therapeutic outcomes for liver cancer by integrating recent progress with remaining challenges.

关键词： Liver tumor detection liver tumor segmentation image processing liver tumor diagnosis feature extraction tumor classification deep learning machine learning

来源：评论

学校读者我要写书评

暂无评论

Assessing classification approaches for categorizing Ayurvedic herbs

引用

Multimedia Tools and Applications 2025年第9期84卷 6711-6735页

作者： Umadevi, v. Kamath, vaishnavi Department of Computer Science and Engineering B.M.S. College of Engineering Bull Temple Road Karnataka Bangalore560019 India

Ayurvedic herbs hold immense importance in traditional Ayurvedic medicine. These herbs are considered vital elements in maintaining physical, mental, and spiritual balance in the human body. According to ayurveda, diseases arise due to imbalances in doshas namely vata, Pitta, and Kapha. Ayurvedic practitioners diagnose diseases by analyzing the dosha imbalances in an individual’s body and mind. Ayurvedic herbs play a crucial role in balancing the doshas and promoting overall health and well-being. Each herb possesses specific properties that can either pacify or aggravate the three *** research employs machine learning models to classify Ayurvedic herbs according to their dosha balancing properties. Using a 90-10 split, the data was randomly split into a training set and a test set. Data preprocessing using random oversampling and artificial sample generation using CTGAN is also performed on the dataset. A total of six experiments are performed on the dataset. various classification models like svM, KNN, RF, DT and XGboost were used for classification. Using four target classes without oversampling, the XGboost classifier yielded the best classification models with an impressive accuracy of 96%, Precision of 97%, Recall of 96%, F1-score of 96% and Root mean square error of 0.16. The experimental results provide compelling evidence that the proposed model, which uses ensemble learning methods, significantly outperforms conventional methods. © The Author(s), under exclusive licence to springer science+business Media, LLC, part of springer Nature 2024.

关键词： Learning systems

来源：评论

学校读者我要写书评

暂无评论

Assessing the predictive power of boosting techniques for diabetes

引用

Multimedia Tools and Applications 2025年 1-25页

作者： v, Umadevi G A, Yashaswini Department of Computer Science and Engineering B.M.S. College of Engineering Bull Temple Road Karnataka Bangalore560019 India

Diabetes affects more than half a million individuals globally. Diabetes can cause heart attacks, renal damage, and even blindness. Machine learning algorithms, with their intelligence capacity can precisely anticipate diabetes and prevent its complications which will allow premature detection of diabetes. boosting techniques offer a powerful approach to machine learning by combining weak models to form a strong ensemble. They excel in addressing challenging classification tasks and can significantly improve predictive performance compared to individual models. This study investigates the predictive efficacy of two widely recognized boosting techniques, namely Adaboost (Ab) and Gradientboost (Gb), across four distinct diabetes datasets. The selected datasets include Pima diabetes dataset (PDD), Early stage diabetes dataset (EDD), Mendeley Diabetes dataset (MDD), and bFRss Diabetes dataset (bDD), to facilitate comprehensive analysis and investigation for diabetes classification. Among the two boosting techniques, the fine tuned Gb showed highest performance with respect to Accuracy, Precision, Recall, F1-score, AUC score, MsE (Mean squared Error), G-Mean (Geometric Mean), Kappa, MCC (Mathews Correlation Coefficient) for diabetes classification on all four diabetes datasets. The accuracy of prediction on four datasets obtained by Gb was 87.66%, 98.07%, 99.00% and 85% respectively. © The Author(s), under exclusive licence to springer science+business Media, LLC, part of springer Nature 2025.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Adaptive Rooftop solar system with Optical Wireless Charging for Electric vehicle batteries 3

Adaptive Rooftop Solar System with Optical Wireless Charging...

引用

3rd International Conference on Electronics and Renewable systems, ICEARs 2025

作者： balamani, T. Manikandan, M. Arthi, b. Logarasu, R. Kumarasamy College Of Engineering Karur India Vel Tech Rangarajan Dr Sagunthala R&d Institute of Science and Technology Chennai India V.S.B Engineering College Karur India Paavai Engineering College Namakkal India

ISBN: (纸本)9798331509675

This study proposes and assesses an adjustable optic wireless power transmission (OWPT) method for charging electric vehicles (Evs) on the ground. This concept proposes to locate energy storage devices, renewable energy resources, and laser transmitters in an overhead facilityAerial Evs can be wirelessly charged using laser transmitter on the building's roof that point upward, and ground Evs can be wirelessly charged using extra laser transmitter on the top floor that point downward. This idea suggests putting laser transmitters, energy storage devices, and renewable energy sources in an above structure. In order for wireless charging of aerial Evs, the building has laser transmitters on its roof that point upward. Additionally, there are other laser transmitters on its ceiling that aim downwards to wireless charge ground Evs. acquired from the roof-top photovoltaic panel Pv based on new design construction. This work presents an Ev charging system that makes use of solar energy from car roofs and is backed by a battery energy storage (bEs) system for parking spaces and mobility. Finally, the implications of the results for the design of ground-based electric vehicles are examined. Wireless power transmission technology indicates that the suggested dynamical OWPT system is more effective © 2025 IEEE.

关键词： Inductive power transmission

来源：评论

学校读者我要写书评

暂无评论

AI Guarded Code Editor: verifying Authentic Human Input

AI Guarded Code Editor: Verifying Authentic Human Input

引用

2025 International Conference on Computing for sustainability and Intelligent Future, COMP-sIF 2025

作者： Raghunandan, Aditi Aishwarya, A.G. Khan, Mohammad Adnan Nishant, s. Panimozhi, K. B.M.S. College of Engineering Department of Computer Science and Engineering Bangalore India

ISBN: (纸本)9798331538538

AI Tools such as ChatGPT, GPT-4 and GitHub Copilot change the very nature of how developers build software by greatly helping in the quick generation of code. While this proves to be efficient for professionals, it poses challenges in the competitive programming space. Generative AI is greatly misused by candidates in coding tests and online interviews. We will design a guarded code editor that detects the probability of usage of GenAI in code submissions with specialized AI detection models that use unique metrics, such as long AI-generated keywords, unusual number of utility functions, and very verbose comments that have been shown to be associated with usual AI-generated content. We will focus on stylistic and structural markers such as redundant comments, too general or too redundant variable names, and predictable formatting to realize a lightweight and scalable detection framework that fits into a user-friendly code editor. Our system also includes mechanisms for immediate feedback that inform the users on the reasons for differences in AI-generated and human-written code, thereby promoting better integrity of code in their own practices. This solution addresses a few critical concerns identified in some recent work that detection tools have limitations to the availability of responding to a domain-specific method responsive to how it would likely be combated using misapplication of the AI system. In other words, with a code editor that does not accept the direct submission of AI-generated code, the proposed solution will most definitely prove effective in both bringing forward practicality and ethical means while ensuring originality and improvement of quality in academic and competitive programming. © 2025 IEEE.

关键词： File editors

来源：评论

学校读者我要写书评

暂无评论

Convolutional-Neural Network bases speaker Emotion Detection based on Human-computer Interaction 1

Convolutional-Neural Network Bases Speaker Emotion Detection...

引用

1st International Conference on Advances in computer science, Electrical, Electronics, and Communication Technologies, CE2CT 2025

作者： Nithya, b. saranya, R. Joseph, Mary Josphineleela, R. Thirunagalingam, Arunkumar Kodali, Arun Kumar B.S. Abdur Rahman Crescent Institute of Science and Technology Dept of Computer Science and Engineering Chennai India V.S.B. Engineering College Department of Computer Science and Engineering Karur India St. Joseph's Institute of Technology OMR Department of IT Chennai119 India Panimalar Engineering College Department of Computer Science and Engineering India Mckesson Corporation United States Mphasis Corporation

ISBN: (纸本)9798331518578

The present investigation proposes a Convolutional Neural Network (CNN)-based method for human-computer interaction (HCI) speaker emotion identification. Optimising the user experience through clearer and sympathetic connections is made possible by emotional knowledge. Conventional techniques for detecting emotions depend on manually created characteristics, which are frequently laborious and less resilient to speech variability. by using CNNs' automatic feature extraction capabilities, the suggested system analyses speech signal spectrograms and extracts crucial emotional indicators. The method obtains notable gains in emotion classification accuracy by retraining the system on a varied collection of emotional speech samples. Important emotional states like happiness, anger, sadness, and neutrality are all successfully distinguished using the CNN model. This study demonstrates how deep learning can help develop emotion-sensitive HCI systems and promote more organic and contextually aware human-machine interactions. Applications in customer service, healthcare, and virtual assistants can benefit from the technique's adaptability and versatility. © 2025 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Multivariate Time series Forecasting for Cloud Machine Utilization 3

Multivariate Time series Forecasting for Cloud Machine Utili...

引用

3rd International Conference on Intelligent Data Communication Technologies and Internet of Things, IDCIoT 2025

作者： Naik, Archana sooda, Kavitha Department of Computer Science and Engineering Nitte Meenakshi Institute of Technology Bangalore India Department of Computer Science and Engineering B.M.S. College of Engineering Bangalore India

ISBN: (纸本)9798331527549

The cloud machine utilization prediction using a multivariate time series model is explored in this study. The cloud machine usage data set samples are selected with varied patterns. The long sequence forecasting model Informer is investigated for the forecasting. The model results are benchmarked against LsTM, GRU, Prophet, and XGboost across the diverse datasets. The performance is evaluated using RMsE. Experimental results show that the Prophet model performs best overall due to its ability to capture both seasonal and irregular patterns accurately. The RMsE value for Informer is 0.13, Prophet is 0.097, XGboost is 0.096, LsTM is 0.1725, and GRU is 0.1547. © 2025 IEEE.

关键词： Time series

来源：评论

学校读者我要写书评

暂无评论

Empirical techniques for effort estimation in designing effective ML models

引用

International Journal of computers and Applications 2025年第2期47卷 226-234页

作者： Padmanabha Reddy, Y.C.A. Kosuru, sai sathwik sirisalla, Nageswara Rao vivekananda, G.N. Perala, venkata Akanksha Department of Artificial Intelligence and Machine Learning Chaitanya Bharati Institute of Technology Telangana Hyderabad India Department of CSE B V Raju Institute of Technology Telangana Narsapur India Department of Computer Science and Engineering K.S.R.M. College of Engineering AP Kadapa India School of Computer Science Engineering and Information Systems Vellore Institute of Technology Vellore India

Machine Learning Research often involves the use of diverse libraries, modules, and pseudocodes for data processing, cleaning, filtering, pattern recognition, and computer intelligence. Quantization of Effort Required for the above cumulative processes is rarely discussed in the existing works and The time to reach the desired level of the model's functionality is essential to gauge the environment training for the model training and pre-deployment testing. In this study, we empirically defined the manual-cum-computational effort required for the model development in terms of 2-time factors: time-to-live (time until 1st code modification) and time-for-modification (time between 2 codebase changes) taken together in a mathematical model to compute Effort Factor(quantitative measure of determining effort needed to design ML algorithms). The study is novel in terms of how the effort required to create ML models is calculated with a particular focus on the manual effort direction. The results and the findings obtained can be used for determining the total time needed to synthesize ML models and frameworks in terms of code change cycles and implementation strategies marking a 25% performance increase in the current method above the standard Pipeline. © 2025 Informa UK Limited, trading as Taylor & Francis Group.

关键词： Pipeline codes

来源：评论

学校读者我要写书评

暂无评论

Toward Privacy-Preserving spatial Crowdsourcing: From Offline to Online

引用

IEEE Network 2025年第3期39卷 326-332页

作者： Wang, Hengzhi Mai, Junjie Zhang, Lei shen, Linfeng Cui, Laizhong Liu, Jiangchuan Shenzhen University College of Computer Science and Software Engineering Shenzhen518060 China Simon Fraser University School of Computing Science BurnabyBCV5A 1S6 Canada

Recent years have seen spatial Crowdsourcing (sC) becoming a promising paradigm of collecting large amounts of location-aware data for various network applications. Its strengths lie in leveraging the power of the crowd to gather data effectively at a low cost. However, during the interaction process between sC participants (e.g., workers and task requestors) and the sC platform, participants’ sensitive information is inevitably exposed, raising severe privacy concerns. Existing studies addressing these privacy concerns have mostly focused on offline sC settings, where the participants’ information is known and accessible in advance. Yet real-world sC participants can be highly dynamic with uncertain behaviors;hence, in practice, their information is largely unknown a priori. In this article, we closely examine the privacy challenges as well as the trade-off between privacy and efficiency in online sC settings. We develop a comprehensive online privacy-preserving sC framework and analyze the critical implementation issues. © 1986-2012 IEEE.

关键词： Crowdsourcing

来源：评论

学校读者我要写书评

暂无评论

Comparative Analysis on Prediction Model of Hate speech Detection based on social Media Data Using Artificial Intelligence 4

Comparative Analysis on Prediction Model of Hate Speech Dete...

引用

4th International Conference on Power, Control and Computing Technologies, ICPC2T 2025

作者： Kaware, Pravin D. Raut, Anjali b. Department of Computer Science & Engineering H.V.P.M's College of Engineering & Technology Amravati India

ISBN: (纸本)9798331530402

The exponential expansion of social media networking platforms has magnified the dissemination of hate speech, thereby requiring strong recognition systems to minimize its influence. The present work offers a comparative examination of many prediction models for the recognition of hate speech using social media data, employing Artificial Intelligence (AI) methodologies. The paper intends to assess the efficiency of several Machine Learning (ML) and Deep Learning (DL) algorithms for accurately detecting hate speech. The various strategies are evaluated using known datasets and measures including accuracy, recall, precision, and Fl-score. In addition, useful insights on the most effective methods for identifying hate speech are presented. The findings suggest that deep learning approaches are exceptionally effective at catching the contextual nuances of hate speech. Al-based hate speech detection apps improve accuracy and efficiency. such systems are critical for creating a safer and more inclusive online. © 2025 IEEE.

关键词： Deep learning

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：