检索结果-内蒙古大学图书馆

Advancing Bankruptcy Forecasting With Hybrid Machine Learning Techniques: Insights From an Unbalanced Polish Dataset

引用

IEEE Access 2024年 12卷 9369-9381页

作者： Ainan, Ummey Hany Por, Lip Yee Chen, Yen-Lin Yang, Jing Ku, Chin Soon Universiti Malaya Faculty of Computer Science and Information Technology Department of Computer System and Technology Kuala Lumpur50603 Malaysia National Taipei University of Technology Department of Computer Science and Information Engineering Taipei106344 Taiwan Universiti Tunku Abdul Rahman Department of Computer Science Kampar31900 Malaysia

The challenge of bankruptcy prediction, critical for averting financial sector losses, is amplified by the prevalence of imbalanced datasets, which often skew prediction models. Addressing this, our study introduces the innovative hybrid model XGBoost+ANN, designed to leverage the strengths of both ensemble learning and artificial neural networks. This model integrates a comprehensive set of features with parameters optimized through genetic algorithms, eschewing traditional feature selection approaches. Our research focuses on an unbalanced dataset of Polish companies and reveals that the XGBoost+ANN model, in particular, exhibits outstanding performance. Optimized using genetic algorithms and without feature selection, this model achieved the highest AUC (0.958), sensitivity (0.752), and accuracy (0.983) scores, surpassing other models in our study. This remarkable outperformance, along with the robust results, marks a substantial advancement in the field of bankruptcy prediction. It underscores the efficacy of our approach in addressing the persistent challenge of data imbalance, offering a more reliable and accurate solution for financial risk assessment. © 2013 IEEE.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Hepatitis C Prediction Using Machine Learning and Deep Learning-Based Hybrid Approach with Biomarker and Clinical Data

引用

Biomedical Materials and Devices 2025年第1期3卷 558-575页

作者： Ripa, Rokiya Uddin, Khandaker Mohammad Mohi Alam, Mir Jafikul Rahman, Md. Mahbubur Department of Computer Science and Engineering Dhaka International University Dhaka 1205 Bangladesh Department of Computer Science and Engineering Southeast University Dhaka 1208 China Department of Computer Science and Engineering Bangladesh University of Business and Technology Dhaka 1216 Bangladesh

Chronic liver damage is believed to be mostly caused by the Hepatitis C virus (HCV). About 90% of hepatitis C infections progress to chronic hepatitis. Acute HCV infection is a condition that frequently progresses to liver cirrhosis and eventually liver cancer;therefore, understanding this stage of the virus is essential. Molecular and serological testing approaches are often expensive and difficult to perform for diagnosing HCV infection. Machine learning technology can be effectively employed to identify patterns or associations for diagnosing HCV infection. The study utilized machine learning techniques to develop classification models for hepatitis C illness, aiming to anticipate the virus responsible for the infection. Our research integrates various machine learning algorithms, including Random Forest, Cat Boost, Bagging Classifier, SGD Classifier, Gaussian NB, Bernoulli NB, Multinomial NB, Linear Discriminant Analysis, ANN, and MLP. Prior to training the models, preprocessing methods, including normalization, filtering, and SMOTE, were used to improve the dataset’s attributes. The classification accuracy scores show encouraging results, with ANN scoring 79.63%, MLP scoring 46.29%, Bernoulli NB scoring 79.62%, Cat Boost scoring 98.14%, and Random Forest scoring 99.53%. Among classification methods, random forest demonstrates the highest accuracy in diagnosing HCV infection. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Classification Deep learning Hepatitis C virus Machine learning Predictions

来源：评论

学校读者我要写书评

暂无评论

KidneyMultiNet: A Web-Based Automatic System for Kidney Disease Detection Using Hybrid Machine Learning Model From CT Scan Images

引用

Biomedical Materials and Devices 2024年 1-12页

作者： Rana, Sahab Uddin Nur-A-Alam, Md. Akter, Sadeka Likhon, Md. Nur Hosain Department of Computer Science and Engineering Dhaka University of Engineering & amp Technology Gazipur Dhaka Bangladesh Department of Computer Science and Engineering Sunamgonj Science and Technology University Sylhet Bangladesh Department of Computer Science and Engineering Dhaka International University Dhaka Bangladesh

Chronic renal disease is the term used to describe kidney function that gradually declines. The kidneys’ final byproduct of eliminating waste and surplus fluid from the bloodstream is urine. Abnormal accumulations of fluid, electrolytes, and waste products can occur in the body when chronic renal disease progresses to a critical degree. Nonetheless, kidney disease has been predicted using various artificial machine learning algorithm approaches. In this study, authors present the “KidneyMultiNet” framework, which combines two models of convolutional neural networks Densenet201 and Xception, which are based on the idea of transfer learning and can predict kidney disease based on kidney CT scan datasets. The “KidneyMultiNet” design that has been suggested makes a thorough and reasonable diagnosis of kidney disease in the kidney conceivable. The authors used a publicly available kidney CT scan image dataset to train and test the proposed model. To determine the efficacy of the output from suggested advanced machine learning CNN methods, the author fed them picture data using a train generator. Then, the author used a rate of learning to decrease the datasets within the proposed model. After completing various learning procedures with many kidney pictures, the proposed model could accurately predict kidney disease, reaching a maximum performance of 99.92%. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Convolutional neural network Image processing Kidney disease Machine learning

来源：评论

学校读者我要写书评

暂无评论

A Generic, High-Performance, Compression-Aware Framework for Data Parallel DNN Training

引用

IEEE Transactions on Parallel and Distributed Systems 2023年 1-20页

作者： Wu, Hao Wang, Shiyi Bai, Youhui Li, Cheng Zhou, Quan Yi, Jun Yan, Feng Chen, Ruichuan Xu, Yinlong Department of Computer Science and Technology University of Science and Technology of China Hefei China Computer Science Department and Electrical and Computer Engineering Department University of Houston USA Nokia Bell Labs

Gradient compression is a promising approach to alleviating the communication bottleneck in data parallel deep neural network (DNN) training by significantly reducing the data volume of gradients for synchronization. While gradient compression is being actively adopted by the industry (e.g., Facebook and AWS), our study reveals that there are two critical but often overlooked challenges: 1) inefficient coordination between compression and communication during gradient synchronization incurs substantial overheads, and 2) developing, optimizing, and integrating gradient compression algorithms into DNN systems imposes heavy burdens on DNN practitioners, and ad-hoc compression implementations often yield surprisingly poor system performance. In this paper, we propose a compression-aware gradient synchronization architecture, CaSync, which relies on flexible composition of basic computing and communication primitives. It is general and compatible with any gradient compression algorithms and gradient synchronization strategies and enables high-performance computation-communication pipelining. We further introduce a gradient compression toolkit, CompLL, to enable efficient development and automated integration of on-GPU compression algorithms into DNN systems with little programming burden. Lastly, we build a compression-aware DNN training framework HiPress with CaSync and CompLL. HiPress is open-sourced and runs on mainstream DNN systems such as MXNet, TensorFlow, and PyTorch. Evaluation via a 16-node cluster with 128 NVIDIA V100 GPUs and a 100 Gbps network shows that HiPress improves the training speed over current compression-enabled systems (e.g., BytePS-onebit, Ring-DGC and PyTorch-PowerSGD) by 9.8%-69.5% across six popular DNN models. IEEE

关键词： Synchronization

来源：评论

学校读者我要写书评

暂无评论

A MobileUnet++-based Abnormality Segmentation and Multi-scale Network Approach for Brain Tumor Classification

A MobileUnet++-based Abnormality Segmentation and Multi-scal...

引用

2024 International Conference on Intelligent Algorithms for Computational Intelligence Systems, IACIS 2024

作者： Kumar, Veesam Pavan Pattanaik, Satya Ranjan Kumar, V Sunil Dept of Computer Science and Engineering Rourkela India Dept of Computer Science and Engineering Bhubaneswar India Pbr Visvodaya Institute of Technology & Science Dept of Computer Science and Engineering Kavali India

ISBN: (纸本)9798350360660

Tumors are a pervasive concern in modern life, driven by cellular irregularities that disrupt the orderly division necessary for healthy cell growth. However, brain tumors present unique challenges compared to tumors in other regions due to the intricacies of brain structure and function. A brain tumor occurs when cells in the cerebral cortex undergo uncontrolled proliferation. Diagnosing brain tumors, especially in their early stages is a formidable task that often relies on manual examination of medical images by radiologists or healthcare professionals. Despite various approaches to brain tumor classification, precision remains paramount while ensuring practical utility in segmentation and classification tasks. Traditional methods are often manual, time-consuming and lacking in automated classification capabilities, hampering efficient decision-making. The advent of deep learning technology has revolutionized medical image processing, particularly in the realm of brain tumor analysis. Deep learning algorithms, powered by vast datasets and computational prowess, offer remarkable capabilities in automating segmentation and classification tasks, thereby enhancing accuracy and efficiency in diagnosing brain tumors. Consequently, this study introduces a method for smart brain tumor segmentation and classification leveraging deep learning techniques. The process begins with acquiring essential images through established benchmarking techniques. These images are then processed in the abnormality segmentation phase, utilizing Mobile-Unet++ (MUnet++) for segmentation. Following segmentation, the images undergo a classification phase, where brain tumors are classified using Multi-scale Mobilenet (MMNet). To validate effectiveness of proposed technique, an efficacy study is conducted, comparing it against traditional methodologies for segmentation and classification of brain. © 2024 IEEE.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Deep belief bi-directional LSTM network-based intelligent student’s performance prediction model with entropy weighted fuzzy rough set mining

引用

International Journal of Intelligent Information and Database Systems 2023年第2期16卷 107-142页

作者： Sateesh, Nayani Rao, P. Srinivasa Lakshmi, D. Rajya Department of Computer Science and Engineering Jawaharlal Nehru Technological University Kakinada Kakinada India Department of Computer Science and Engineering MVGR College of Engineering Vizianagaram India Department of Computer Science and Engineering University College of Engineering Vizianagaram Jawaharlal Nehru Technological University Kakinada India

To analyse the student’s academic performance, a new prediction model is developed. This proposed model collects the student’s data from standard online sources. At first, these gathered data are pre-processed by certain methods like outlier removal and filling. Then, the filtered data undergoes feature mining by developing the entropy weighted fuzzy rough set mining (EW-FRSM), where certain parameters are tuned with the proposed hybrid optimisation algorithm with probability-based coyote electric fish optimisation (P-CEFO). The refined features are utilised for performance prediction, where the deep belief-Bi-LSTM (DB-Bi-LSTM) is used for student performance prediction. The parameters in DBN and Bi-LSTM are tuned with the same P-CEFO algorithm to improve the prediction performance. Simulation is performed to examine the efficacy of the designed prediction method with diverse performance metrics by evaluating with diverse baseline approaches to prove the usefulness of the model. Copyright © 2023 Inderscience Enterprises Ltd.

关键词： Entropy

来源：评论

学校读者我要写书评

暂无评论

Powder spreading process monitoring of selective laser melting manufacturing by using a convolutional Takagi–Sugeno–Kang fuzzy neural network

引用

International Journal of Advanced Manufacturing Technology 2024年第9-10期132卷 4989-5004页

作者： Lin, Chun-Hui Lin, Cheng-Jian Wang, Shyh-Hau Department of Computer Science and Information Engineering National Cheng Kung University Tainan701 Taiwan Department of Computer Science and Information Engineering National Chin-Yi University of Technology Taichung411 Taiwan

Process monitoring plays a pivotal role in elucidating the intricate interplay among process, structure, and property in additive manufacturing production. The control of powder spreading affects not only particle adhesion but also rearrangement, both of which are indispensable constituents in the selective laser melting (SLM) process. To develop a novel monitoring process, this study proposed a convolutional Takagi–Sugeno–Kang fuzzy neural network (C-TSKFNN) to assess images with uneven powder distribution. The performance of the model was compared with that of conventional image analysis and state-of-the-art deep learning models. With its fusion layer and TSKFNN design, the C-TSKFNN model retains the distinctive characteristics of input images and minimizes expenditures through trainable parameters. The experimental results demonstrated that the average F1 score of the C-TSKFNN model achieved 97.07% and that it had the lowest standard deviation compared to state-of-the-art deep learning models. The trainable parameters were only 103,506, which is lower than those of models. Defect information was also provided locally through a model interpretation technique, facilitating the monitoring of SLM manufacturing. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Process monitoring

来源：评论

学校读者我要写书评

暂无评论

Automatic Sentimental Analysis by Firefly with Levy and Multilayer Perceptron

引用

computer Systems science & engineering 2023年第9期46卷 2797-2808页

作者： D.Elangovan V.Subedha School of Computing Computer Science and EngineeringSathyabama Institute of Science and TechnologyChennai600118India Computer Science and Engineering Panimalar Institute of TechnologyChennai600069India

The field of sentiment analysis(SA)has grown in tandem with the aid of social networking platforms to exchange opinions and *** people share their views and ideas around the world through social media like Facebook and *** goal of opinion mining,commonly referred to as sentiment analysis,is to categorise and forecast a target’s *** on if they provide a positive or negative perspective on a given topic,text documents or sentences can be *** compared to sentiment analysis,text categorization may appear to be a simple process,but number of challenges have prompted numerous studies in this area.A feature selection-based classification algorithm in conjunction with the firefly with levy and multilayer perceptron(MLP)techniques has been proposed as a way to automate sentiment analysis(SA).In this study,online product reviews can be enhanced by integrating classification and feature *** firefly(FF)algorithm was used to extract features from online product reviews,and a multi-layer perceptron was used to classify sentiment(MLP).The experiment employs two datasets,and the results are assessed using a variety of *** account of these tests,it is possible to conclude that the FFL-MLP algorithm has the better classification performance for Canon(98%accuracy)and iPod(99%accuracy).

关键词： Firefly algorithm feature selection feature extraction multi-layer perceptron automatic sentiment analysis

来源：评论

学校读者我要写书评

暂无评论

Unsupervised Domain Adaptation on Point Clouds via High-Order Geometric Structure Modeling

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2024年第12期5卷 6121-6133页

作者： Cheng, Jiang-Xing Lin, Huibin Zhang, Chun-Yang Chen, C. L. Philip Fuzhou University College of Computer and Data Science Fuzhou350108 China South China University of Technology School of Computer Science and Engineering Guangdong Guangzhou510006 China

Point clouds can capture the precise geometric information of objects and scenes, which are an important source of 3-D data and one of the most popular 3-D geometric data structures for cognitions in many real-world applications like automatic driving and remote sensing. However, due to the influence of sensors and varieties of objects, the point clouds obtained by different devices may suffer obvious geometric changes, resulting in domain gaps that are prone to the neural networks trained in one domain failing to preserve the performance in other domains. To alleviate the above problem, this article proposes an unsupervised domain adaptation framework, named HO-GSM, as the first attempt to model high-order geometric structures of point clouds. First, we construct multiple self-supervised tasks to learn the invariant semantic and geometric features of the source and target domains, especially to capture the feature invariance of high-order geometric structures of point clouds. Second, the discriminative feature space of target domain is acquired by using contrastive learning to refine domain alignment to specific class level. Experiments on the PointDA-10 and GraspNetPC-10 collection of datasets show that the proposed HO-GSM can significantly outperform the state-of-the-art counterparts. © 2020 IEEE.

关键词： Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

An efficient security testing for android application based on behavior and activities using RFE-MLP and ensemble classifier

引用

Multimedia Tools and Applications 2025年第15期84卷 15357-15386页

作者： Kumar, Pawan Singh, Sukhdip Department of Computer Science and Engineering Deenbandhu Chhotu Ram University of Science and Technology Murthal Sonipat131039 India

An enormous amount of applications that are available for download permits users to enhance the functionality of the devices with brand-new features, which is a significant factor in the growing popularity of smartphones and tablets. In reality, third-party applications for business and entertainment are abundantly available in official stores, with the majority of them being offered for free. However, vulnerable applications may expose private information (such as emails, global GPS location, financial information, and phone contacts). To overcome these issues, the log files are collected for normalizing and selecting the features related to the behaviour and activities using the Min–Max and Recursive Feature Elimination (RFE) based on the MLP. The selected features are learned by using the ensemble learning algorithm, which is a stacking of Bi-directional Long Short Term Memory (Bi-LSTM), Deep Neural Network (DBN) and Radial Basis Functional Network (RBFN) giving the collected learned features from each algorithm to the Meta classifier for predicting the malware. The ensemble learning algorithm then feeds the gathered learned features from each method to the Meta classifier for malware prediction. The suggested strategy protects users' privacy by alerting them when it detects malicious apps and removing them. Accuracy, Precision, Specificity, and Kappa performance measures are assessed and contrasted with the proposed and current models. The proposed model achieved performance metrics values are 99.78%, 99.56%, 99.85% and 98.89%. Thus, the security testing for Android applications based on behaviour and activities using RFE-MLP and Ensemble classifier performs better than the existing model. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Long short-term memory

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：