检索结果-内蒙古大学图书馆

EGRNN++ and PNN++ : Parallel and Distributed Neural Networks for big data Regression and classification

SN Computer Science 2021年第2期2卷 109页

作者： Kamaruddin, Sk Ravi, Vadlamani Centre of Excellence in Analytics Institute for Development and Research in Banking Technology Castle Hills Road No. 1 Masab Tank Hyderabad 500057 India SCIS University of Hyderabad Hyderabad 500046 India

Probabilistic Neural Network (PNN) and General Regression Neural Network (GRNN) are unique as they are trained in one pass and perform well for classification and regression problems, respectively;however, they cannot handle big data sets, because the pattern layer in them is required to store all the training samples. Therefore, this paper proposes hybrid architectures for PNN and GRNN, where the pattern layer is made simpler by storing cluster centers of all the samples, thereby making them amenable for big data analytics. This paper proposes the use of parallel distributed clustering algorithms viz., K-Means|| and parallel Bisecting K-Means before the pattern layer of GRNN and PNN. We proposed the use of logistic and Cauchy activation functions in the place of Gaussian activation function for both networks. The effectiveness of the variants of EGRNN++ was tested on two data sets taken from Chemistry and Amazon Movie Review dataset for customer review rating prediction under a ten-fold cross-validation (10-FCV) setup. Similarly, the performance of the variants of PNN++ was measured with HEPMASS, HIGGS, ccFraud and Amazon Movie Review dataset for sentiment classification under a 10-FCV setup. The proposed variants of EGRNN++ produced very low Mean Squared Error (MSE). The proposed variants of PNN++ produced high AUC. We have conducted Wilcoxon Signed-Rank test at 1% level of significance. It is worthwhile to mention that the primary objective of this article is to present a distributed and parallel version of the traditional GRNN and PNN to handle big data sets. © 2021, The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd. part of Springer Nature.

关键词： Apache spark big data classification big data regression Bisecting K-means|| GRNN K-means|| PNN Sentiment analysis

来源：评论

学校读者我要写书评

暂无评论

A hybrid multi-objective firefly and simulated annealing based algorithm for big data classification

引用

CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE 2019年第14期31卷 e4985-e4985页

作者： Devi, S. Gayathri Sabrigiriraj, M. Coimbatore Inst Engn & Technol Dept Informat Technol Coimbatore 641109 Tamil Nadu India SVS Coll Engn Dept Elect & Commun Engn Coimbatore Tamil Nadu India

Efficient management of big data becomes challenging in recent decades. Online Feature Selection (OFS) is one type of online learning in contrast to batch learning, allowing a classifier to have small and fixed number of features. The major aim of this work is to introduce an OFS algorithm supported on meta-heuristic algorithm that exploits the MapReduce paradigm. A novel Hybrid Multi-Objective Firefly and Simulated Annealing (HMOFSA) algorithm is proposed to select optimal set of features. Therefore, as a first step, the original big dataset is decomposed into blocks of examples in the map phase. Subsequently, HMOFSA algorithm is employed to choose the selected features from examples. After that, the attained partial outcomes will be combined into a final vector of features in the reduce phase and evaluated using Kernel Support Vector Machine (KSVM) classifier. The mentioned OFS approach is analyzed with the help of the well-known classifiers (Logistic Regression, KSVM and Naive Bayes) developed within the Spark framework. Experiments were conducted on big datasets, containing 66 million samples and 2000 attributes that confirm the proficiency of proposed work. The proposed KSVM classifier results are measured in terms of the metrics like Precision, Recall, Geometric-mean (G-mean), F-measure, and accuracy.

关键词： big data big data classification Hybrid Multi-Objective Firefly and Simulated Annealing (HMOFSA) algorithm Kernel Support Vector Machine (KSVM) MapReduce(MR) paradigm meta-heuristic algorithm Online Feature Selection (OFS)

来源：评论

学校读者我要写书评

暂无评论

Feature selection based on an improved cat swarm optimization algorithm for big data classification

引用

JOURNAL OF SUPERCOMPUTING 2016年第8期72卷 3210-3221页

作者： Lin, Kuan-Cheng Zhang, Kai-Yuan Huang, Yi-Hung Hung, Jason C. Yen, Neil Natl Chung Hsing Univ Dept Management Informat Syst Taichung 40227 Taiwan Natl Taichung Univ Educ Dept Math Educ Taichung Taiwan Overseas Chinese Univ Dept Informat Technol Taichung Taiwan Univ Aizu Sch Comp Sci & Engn Aizu Wakamatsu Fukushima Japan

Feature selection, which is a type of optimization problem, is generally achieved by combining an optimization algorithm with a classifier. Genetic algorithms and particle swarm optimization (PSO) are two commonly used optimal algorithms. Recently, cat swarm optimization (CSO) has been proposed and demonstrated to outperform PSO. However, CSO is limited by long computation times. In this paper, we modify CSO to present an improved algorithm, ICSO. We then apply the ICSO algorithm to select features in a text classification experiment for big data. Results show that the proposed ICSO outperforms traditional CSO. For big data classification, the results show that using term frequency-inverse document frequency (TF-IDF) with ICSO for feature selection is more accurate than using TF-IDF alone.

关键词： Cat swarm optimization Feature selection Support vector machine big data classification

来源：评论

学校读者我要写书评

暂无评论

Elastic extreme learning machine for big data classification

引用

NEUROCOMPUTING 2015年第PartA期149卷 464-471页

作者： Xin, Junchang Wang, Zhiqiong Qu, Luxuan Wang, Guoren Northeastern Univ Coll Informat Sci & Engn Shenyang Peoples R China Northeastern Univ Sino Dutch Biomed & Informat Engn Sch Shenyang Peoples R China

Extreme Learning Machine (ELM) and its variants have been widely used for many applications due to its fast convergence and good generalization performance. Though the distributed ELM* based on MapReduce framework can handle very large scale training dataset in big data applications, how to cope with its rapidly updating is still a challenging task. Therefore, in this paper, a novel Elastic Extreme Learning Machine based on MapReduce framework, named Elastic ELM ((ELM)-L-2), is proposed to cover the shortage of ELM* whose learning ability is weak to the updated large-scale training dataset. Firstly, after analyzing the property of ELM* adequately, it can be found out that its most computation-expensive part, matrix multiplication, can be incrementally, decrementally and correctionally calculated. Next, the Elastic ELM based on MapReduce framework is developed, which first calculates the intermediate matrix multiplications of the updated training data subset, and then update the matrix multiplications by modifying the old matrix multiplications with the intermediate ones. Then, the corresponding new output weight vector can be obtained with centralized computing using the update the matrix multiplications. Therefore, the efficient learning of rapidly updated massive training dataset can be realized effectively. Finally, we conduct extensive experiments on synthetic data to verify the effectiveness and efficiency of our proposed (ELM)-L-2 in learning massive rapidly updated training dataset with various experimental settings. (C) 2014 Elsevier B.V. All rights reserved.

关键词： Extreme learning machine big data classification Incremental learning Decremental learning Correctional learning

来源：评论

学校读者我要写书评

暂无评论

An Approach to Sample Selection from big data for classification

An Approach to Sample Selection from Big Data for Classifica...

引用

IEEE International Conference on Systems, Man, and Cybernetics (SMC)

作者： Xing, Sheng He, Yulin Zhu, Hong Wang, Xizhao HeBei Univ Coll Management Baoding 071002 Hebei Province Peoples R China Shenzhen Univ Coll Comp Sci & Software Engn Shenzhen 518060 Guangdong Peoples R China Macau Univ Sci & Technol Fac Informat Technol Taipa Macao Special A Peoples R China Shenzhen Univ Coll Comp Sci & Software Engn Shenzhen 518060 Guangdong Peoples R China

ISBN: (纸本)9781509018970

When traditional sample selection methods are used to compress large data sets, the computational complexity turns out to be very high and it is really time consuming. To avoid these shortcomings, we propose a new method to select samples based on non-stable cut points. With the basic characteristic of convex function that its extreme values occur at the endpoints of intervals, the method measures the extent of a sample being endpoints by labeling non-stable cut points. Then we can select the samples with higher endpoint extent, which can avoid calculating the distances between samples. This method aims to compress the data sets and improve the computational efficiency without affecting the classification accuracy. Experiments show that the proposed algorithm performs very well on the compression of data sets with higher imbalance degree. Meanwhile, the method is experimentally confirmed to have strong noise-resistance.

关键词： big data classification Sample selection Non-stable cut points Decision tree

来源：评论

学校读者我要写书评

暂无评论

Fuzzy Neighbors and Deep Learning-Assisted Spark Model for Imbalanced classification of big data

引用

INTERNATIONAL JOURNAL OF UNCERTAINTY FUZZINESS AND KNOWLEDGE-BASED SYSTEMS 2023年第1期31卷 141-162页

作者： Nalinipriya, G. Geetha, M. Sudha, D. Daniya, T. Saveetha Engn Coll Dept Informat Technol Chennai 602105 Tamil Nadu India Chennai Inst Technol Dept Comp Sci & Engn Chennai Tamil Nadu India Meenakshi Coll Engn Dept Comp Sci & Engn Chennai 600078 Tamil Nadu India GMR Inst Technol Dept Informat Technol Rajam 532127 Andhra Prades India

big data is important in knowledge manipulation, assessment, and prediction. However, extracting and analyzing knowledge through big database are complex because of imbalance data distribution that leads to wrong decisions and biased classification outputs. Hence, an effective and optimal big data classification approach is designed using the proposed Bird Swarm Deer Hunting Optimization-Deep Belief Network (BSDHO-based DBN) algorithm based on spark architecture that follows the master and slave nodes. The proposed BSDHO is obtained by combining Deer Hunting Optimization algorithm and Bird Swarm Algorithm. The developed model poses two nodes, namely slave and master node. The training data is initially given to the master node in the spark architecture to perform transformation of data. Here, the transformation of data is done with an exponential log kernel, and then selection of feature is done with sequential forward selecting for choosing suitable features for enhanced processing. Consequently, oversampling process is performed with Fuzzy K-Nearest Neighbor (Fuzzy KNN) in the slave node using selected features to manage imbalance data. Then, in master node, classification is done with Deep belief Network, and trained using developed Bird swarm Deer Hunting Optimization (BSDHO) algorithm. On the other hand, the test data is taken as input, and is fed to the slave node to perform data transformation. Then, the transformed data is given to the master node for classification based on the proposed BSDHO. At last, the training data and testing data output produced the classified output. The proposed BSDHO-based DBN provided enhanced outcomes with highest specificity of 97.92%, accuracy of 96.92%, and sensitivity of 96.9%.

关键词： big data classification spark architecture Bird Swarm optimization Deep Belief network Deer hunting optimization

来源：评论

学校读者我要写书评

暂无评论

big data classification based on multi-view method

Big data classification based on multi-view method

引用

International Conference on Wavelet Analysis and Pattern Recognition (ICWAPR)

作者： Weiwen Liu Danni Chen School of Computer Science and Engineering South China University of Technology Guangzhou China

ISBN: (纸本)9781467372251

The traditional classification technique cannot be applied directly to the big data problem mainly due to the high time complexity issue. The existing method trained a classifier using the dataset simplified by the clustering approach. However, the main drawback of this method is that the dataset may be over- or under-simplified. In this paper, we revises this method for a big data classification task. Our proposed method considers multi-view rather than single training set. Several simplified datasets are firstly generated based on the clustering technique with different radii. Then several classifiers are trained on these datasets and are combined together based on the boosting method to form a multiple classifier system (MCS). Therefore, different levels of the dataset can be learnt, and also the time complexity of the classification method is practicable. The experimental result suggests that our method achieves better performance in comparison to the traditional SVM method on the big data problem.

关键词： big data computational complexity learning (artificial intelligence) pattern classification big data classification MCS boosting method classifier training clustering approach multiple classifier system multiview method over-simplified dataset time complexity under-simplified dataset big data classification Clustering Multiple Classifier System big data complexity classes dataset MCS gene Pattern recognition Cluster Analysis Classifiers

来源：评论

学校读者我要写书评

暂无评论

A MapReduce-based k-Nearest Neighbor Approach for big data classification

A MapReduce-based k-Nearest Neighbor Approach for Big Data C...

引用

IEEE International Conference on big data Science and Engineering

作者： Jesús Maillo Isaac Triguero Francisco Herrera Department of Computer Science and Artificial Intelligence University of Granada 18071 Granada Spain Department of Respiratory Medicine Ghent University 9000 Gent Belgium VIB Inflammation Research Center 9052 Zwijnaarde Belgium

ISBN: (纸本)9781467379533

The k-Nearest Neighbor classifier is one of the most well known methods in data mining because of its effectiveness and simplicity. Due to its way of working, the application of this classifier may be restricted to problems with a certain number of examples, especially, when the runtime matters. However, the classification of large amounts of data is becoming a necessary task in a great number of real-world applications. This topic is known as big data classification, in which standard data mining techniques normally fail to tackle such volume of data. In this contribution we propose a MapReduce-based approach for k-Nearest neighbor classification. This model allows us to simultaneously classify large amounts of unseen cases (test examples) against a big (training) dataset. To do so, the map phase will determine the k-nearest neighbors in different splits of the data. Afterwards, the reduce stage will compute the definitive neighbors from the list obtained in the map phase. The designed model allows the k-Nearest neighbor classifier to scale to datasets of arbitrary size, just by simply adding more computing nodes if necessary. Moreover, this parallel implementation provides the exact classification rate as the original k-NN model. The conducted experiments, using a dataset with up to 1 million instances, show the promising scalability capabilities of the proposed approach.

关键词： k-Nearest Neighbor Approach big data classification simplicity

来源：评论

学校读者我要写书评

暂无评论

A dynamic ensemble learning based data mining framework for medical imbalanced big data

引用

KNOWLEDGE-BASED SYSTEMS 2025年 310卷

作者： Rithani, M. Kumar, R. Prasanna Ali, Altalbe Amrita Vishwa Vidyapeetham Amrita Sch Comp Dept Comp Sci & Engn Chennai India Prince Sattam Bin Abdulaziz Univ Dept Comp Engn AlKharj Saudi Arabia King Abdulaziz Univ Fac Comp & Informat Technol Jeddah Saudi Arabia

In the era of big data, technologies like the Internet of Things, smart cities, healthcare, and social media rely heavily on advanced data analytics. In medical data, certain critical diseases are significantly underrepresented compared to more prevalent conditions, creating a class imbalance that can lead to biased models favoring majority class predictions. This imbalance reduces the accuracy and reliability of predictions for the minority class, which is often essential for early diagnosis and intervention in rare but severe diseases. This is particularly challenging in medical data, where cancer classification faces problems such as high dimensionality, redundancy, and severe class imbalance. To address these challenges, this paper proposes a novel framework which integrates a Relevance Vector Machine classifier with an Incremental Ensemble framework to effectively manage data imbalance. It employs a Gaussian Mixture Models-based combined resampling algorithm to balance the dataset by resampling. Mutual Information Gain Maximization enhances the effectiveness of feature selection. To further enhance performance, an Adaptive Weighted Broad Learning System is incorporated a density-based weight generation mechanism using prior distribution information. Additionally, an Incremental Dynamic Learning Policy-based Relevance Vector Machine classifier is incorporated to adapt to new data, and maintain high accuracy. The proposed model achieves superior performance with an Accuracy of 99 %, a Kappa value of 98 %, an F1-Score of 99 %, and an MCC of 96.9 %. These results underscore the model's effectiveness in addressing class imbalance, enhancing predictive accuracy for minority classes, and offering a robust solution for complex medical datasets essential for improved healthcare outcomes.

关键词： Relevance vector machine Mutual information gain maximization Gaussian mixture model big data classification Incremental weighted ensemble broad learning system

来源：评论

学校读者我要写书评

暂无评论

Storage and Query Indexing Methods on big data

引用

ARABIAN JOURNAL FOR SCIENCE AND ENGINEERING 2024年第5期49卷 7359-7374页

作者： Wu, QingE Yu, Yao Zhou, Lintao Lu, Yingbo Chen, Hu Qian, Xiaoliang Zhengzhou Univ Light Ind Sch Elect & Informat Engn Zhengzhou 450002 Peoples R China

For solving the storage and indexing problem of big data, this paper proposes a big data storage algorithm and an indexing algorithm, respectively. Simultaneously, the data classification and segmentation algorithm is given, and then, the classified data are segmented and sliced based on features of data. Moreover, the calculation of the service queue length on data slicing is given to achieve fast storage of slices in the data server. A data reconstruction method is given to index each slice of the original data and implement the data reconstruction according to the given reconstruction method. At the same time, an evaluation method for data reconstruction is also given to achieve the minimum difference between the reconstructed data and the original data. Simulation results show that the accurate rate of the data reconstruction method proposed in this paper is higher than 93%. In addition, compared with the existing data reconstruction methods, it is found that it not only has higher reconstruction accurate rate, faster reconstruction process, stronger anti-interference ability, but also lower information loss. These researches in this paper provide the powerful methods for protecting important documents, reducing losses, reducing waste of resources, and improving the convenience, speed, and safety in work, life, and so on.

关键词： big data classification Service queue length data reconstruction Evaluation method

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：