检索结果-内蒙古大学图书馆

Rolling Bearing Fault Diagnosis Method Based on Eigenvalue Selection and dimension reduction algorithm

INTERNATIONAL JOURNAL OF PATTERN RECOGNITION AND ARTIFICIAL INTELLIGENCE 2021年第9期35卷 2150027-2150027页

作者： Chen, Song Chen, Li-Ai Wang, Da-Gui Cheng, He-Sheng Anhui Jianzhu Univ Coll Mech & Elect Engn Hefei 230601 Anhui Peoples R China Anhui Prov Key Lab Simulat & Design Elect Informa Hefei 230601 Anhui Peoples R China

The bearings used in the mechanical equipment that bear and transfer the load are vulnerable parts. In this paper, a rolling bearing fault diagnosis method based on eigenvalue selection and dimensionality reduction is presented. This is suitable for analyzing fault signals with nonstationary characteristics, and it has a good recognition rate. The characteristic quantity of vibration signals in the time domain and the frequency domain is calculated, and the characteristic quantity is selected by calculating the degree of difference. A dimension reduction algorithm is proposed, which is based on a neural network and ISOMAP. Its performance is compared using PCA, LTSA, and ISOMAP algorithms. Fault diagnosis is carried out by using KNN and SVM classification algorithms, and good recognition results are obtained.

关键词： Fault diagnosis ISOMAP support vector machines dimension reduction algorithm

来源：评论

学校读者我要写书评

暂无评论

Estimating gene expression from high-dimensional DNA methylation levels in cancer data: A bimodal unsupervised dimension reduction algorithm

引用

COMPUTERS & INDUSTRIAL ENGINEERING 2019年 130卷 348-357页

作者： Damgacioglu, Haluk Celik, Emrah Celik, Nurcin Univ Miami Dept Ind Engn Coral Gables FL 33124 USA Univ Miami Dept Mech & Aerosp Engn Coral Gables FL 33124 USA

Recent molecular and genetic studies have revealed the importance of DNA methylation, a key epigenetic mark, in regulating gene expression and the abnormal profiles of DNA methylation in various diseases including cancer. Here, unsupervised learning methods that are geared towards high-throughput DNA methylation analysis are used to extract useful information from high-dimensional genome wide methylation data in order to provide crucial insights for accurate early diagnosis and treatment of cancer. Herein, these methods are highly dependent on the performance of an earlier step of dimension reduction that aims to find the best subset of attributes to be retained for learning. Widely used algorithms in the literature commonly suffer from resulting in trivial cluster structures and failing to shed light on the relationship between DNA methylation and cancer types due to their myopic and arbitrary search mechanisms. Addressing this issue, we introduce a bimodal unsupervised dimension reduction algorithm (BOUNDER) that identifies the best subset of loci for downstream analysis considering the variability and redundancy across all the samples using bimodal modeling before it feeds into the learning method. BOUNDER models each locus as a bimodal representation using a piecewise linear function with two segments and filters the informative loci based on the fitted line characteristics. To the best of our knowledge, the work presented here is the first study that uses bimodal modeling in unsupervised learning in DNA methylation analysis. BOUNDER is tailored for DNA methylation analysis using a detailed parameter tuning analysis. The performance of BOUNDER is benchmarked against those of widely used conventional algorithms using real lung, breast, kidney, and urological cancer datasets obtained from Gene Expression Omnibus in terms of their accuracies in hierarchical clustering and k-means clustering. Computational experiments reveal that BOUNDER outperforms the PCA and filt

关键词： DNA methylation data Values dimension reduction algorithm Big data analytics Beta distribution Piece linear curve fitting

来源：评论

学校读者我要写书评

暂无评论

Exploring Aromatase Inhibitors to Breast Cancer with Machine Learning Approach Based Quantitative Structure-Activity Relationships Model Technique 7

Exploring Aromatase Inhibitors to Breast Cancer with Machine...

引用

2023 7th International Conference On Computing, Communication, Control And Automation, ICCUBEA 2023

作者： Pawar, Sahebrao B. Deshmukh, N.K. School of Computational Sciences Swami Ramanand Teerth Marathwada University Nanded India

ISBN: (纸本)9798350304268

The conversion of androgen depends in large part on the overexpression of aromatase in many females. Many therapeutic and pharmaceutical drugs, as well as inhibitors, have been developed, but some of them either don't have the strength to control diseases caused by the aromatase enzyme or they encourage unwelcome mammary epithelial cells inside the breast. The prospect and potential for creating medicines with unique pharmacological and therapeutic capabilities using the Multi-Machine Learning Classifier Based-Quantitative Structure-Activity Relationship (MMLCB-QSAR) process model were looked into. In order to forecast the chemical classes with the highest potential for Aromatase, The model uses chemical descriptors and supervised machine learning to inhibit utilizing real multidimensional Aromatase inhibitors that were collected from a curated database. The multi-dimensional dataset is mainly explored and visualized using the t-SNE algorithm. Despite having excellent predictive abilities, the Bootstrap aggregation or Bagging Classifier (RandomForest Classifier) algorithm excels. More specifically, regularization and hyperparameter tuning are responsible for outstanding model results in statistics and more accurate prediction using 10-fold cross-validation. Model parameters like the Matthews correlation coefficient score (0.9397), the accuracy Score (0.973%), and the log loss probability (0.1208). © 2023 IEEE.

关键词： Aromatase Enzyme dimension reduction algorithm MMLCB-Quantitative Structure-Activity Relationships PaDEL-Descriptor.jar Software PubChem Fingerprint QSAR

来源：评论

学校读者我要写书评

暂无评论

Advanced survival prediction in head and neck cancer using hybrid machine learning systems and radiomics features

Advanced survival prediction in head and neck cancer using h...

引用

Conference on Medical Imaging - Biomedical Applications in Molecular, Structural, and Functional Imaging

作者： Salmanpour, Mohammadreza R. Hosseinzadeh, Mahdi Modiri, Ensiyeh Akbari, Azizeh Hajianfar, Ghasem Askari, Dariush Fatan, Mehdi Maghsudi, Mehdi Ghaffari, Hanieh Rezaei, Masoad Ghaemi, Mohammad M. Rahmim, Arman Univ British Columbia Vancouver BC Canada BC Canc Res Inst Vancouver BC Canada Technol Virtual Collaborat TECVICO Corp Vancouver BC Canada Tarbiat Modares Univ Tehran Iran Islamic Azad Univ Tehran Iran Hakim Sabzevari Univ Sabzevar Iran Iran Univ Med Sci Rajaie Cardiovasc Med & Res Ctr Tehran Iran Shahid Beheshti Univ Med Tehran Iran Ahvaz Jundishapur Univ Med Sci Sch Med Dept Med Phys Ahvaz Iran Kerman Univ Med Sci Tehran Iran

ISBN: (数字)9781510649484

ISBN: (纸本)9781510649484;9781510649477

Accurate prognostic stratification of Head-and-Neck-Squamous-Cell-Carcinoma (HNSCC) patients can be an important clinical reference when designing therapeutic strategies. We set to predict 4 outcomes: overall survival (OS), distant metastasis (DM), locoregional recurrence (LR), and progression-free survival (SP). We studied Hybrid Machine Learning Systems (HMLS), applied to datasets with radiomics features. In this multicenter study, 408 HNSCC patients were extracted from The Cancer Imaging Archive (TCIA) database. PET images were registered to CT, enhanced, and cropped. 215 radiomics features were extracted from each region of interest via our standardized SERA radiomics package. We employed multiple HMLSs: 12 feature extraction ( FEA) or 9 feature selection algorithms (FSA) linked with 9 survival-prediction-algorithms (SPA) optimized by 5-fold cross-validation, applied to PET only, CT only and 4 PET-CT datasets generated by image-level fusion strategies. Datasets were normalized by z-score-technique, and c-indices were reported to compare the models. For OS prediction, the highest c-index 0.73 +/- 0.10 was obtained for HMLS with Ratio of low-pass pyramid (RP) fusion technique + gaussian process latent variable model ( GPLVM) + causal structure learning-based feature modification method (CSFM). For DM prediction, we achieved 0.80 +/- 0.06 via Dual-tree complex wavelet transform (DTCWT) fusion + Laplacian Score (LAP) + Logistic regression hazards (LH). For LR prediction, we arrived at a c-index of 0.73 +/- 0.13 using PET + Sammon Mapping algorithm (SM)+ deep neural network to distribute first hitting times (DHS). For SP prediction, the performance of 0.68 +/- 0.02 was obtained via PET + SM + Relative risk model-depend on time (CoxTime). When no dimensionality reduction (FEA/FSA) was employed, the above 4 performances decreased to 0.69 +/- 0. 10, 0.74 +/- 0.13, 0.66 +/- 0.15, and 0.68 +/- 0.04 for OS, DM, LR and SP prediction. We demonstrated that using fusion techni

关键词： survival prediction hybrid machine learning system radiomics features fusion techniques head and neck squamous cell carcinoma image processing dimension reduction algorithm multicenter study

来源：评论

学校读者我要写书评

暂无评论

Attention Mechanism and Depthwise Separable Convolution Aided 3DCNN for Hyperspectral Remote Sensing Image Classification

引用

REMOTE SENSING 2022年第9期14卷 2215-2215页

作者： Li, Wenmei Chen, Huaihuai Liu, Qing Liu, Haiyan Wang, Yu Gui, Guan Nanjing Univ Posts & Telecommun Sch Geog & Biol Informat Nanjing 210023 Peoples R China Nanjing Univ Posts & Telecommun Coll Telecommun & Informat Engn Nanjing 210003 Peoples R China Nanjing Univ Posts & Telecommun Smart Hlth Big Data Anal & Locat Serv Engn Lab Ji Nanjing 210023 Peoples R China

Hyperspectral Remote Rensing Image (HRSI) classification based on Convolution Neural Network (CNN) has become one of the hot topics in the field of remote sensing. However, the high dimensional information and limited training samples are prone to the Hughes phenomenon for hyperspectral remote sensing images. Meanwhile, high-dimensional information processing also consumes significant time and computing power, or the extracted features may not be representative, resulting in unsatisfactory classification efficiency and accuracy. To solve these problems, an attention mechanism and depthwise separable convolution are introduced to the three-dimensional convolutional neural network (3DCNN). Thus, 3DCNN-AM and 3DCNN-AM-DSC are proposed for HRSI classification. Firstly, three hyperspectral datasets (Indian pines, University of Pavia and University of Houston) are used to analyze the patchsize and dataset allocation ratio (Training set: Validation set: Test Set) in the performance of 3DCNN and 3DCNN-AM. Secondly, in order to improve work efficiency, principal component analysis (PCA) and autoencoder (AE) dimension reduction methods are applied to reduce data dimensionality, and maximize the classification accuracy of the 3DCNN, but it will still take time. Furthermore, the HRSI classification model 3DCNN-AM and 3DCNN-AM-DSC are applied to classify with the three classic HRSI datasets. Lastly, the classification accuracy index and time consumption are evaluated. The results indicate that 3DCNN-AM could improve classification accuracy and reduce computing time with the dimension reduction dataset, and the 3DCNN-AM-DSC model can reduce the training time by a maximum of 91.77% without greatly reducing the classification accuracy. The results of the three classic hyperspectral datasets illustrate that 3DCNN-AM-DSC can improve the classification performance and reduce the time required for model training. It may be a new way to tackle hyperspectral datasets in HRSl classificati

关键词： hyperspectral remote sensing image classification attention mechanism depthwise separable convolution three-dimensional convolutional neural network dimension reduction algorithm

来源：评论

学校读者我要写书评

暂无评论

A Multi-Closed-Loop Assessment Mechanism for Development Level of Smart Grids

引用

FRONTIERS IN ENERGY RESEARCH 2022年 10卷

作者： Xie, Haodong Li, Zhenqiu Yang, Qixiong Peng, Zhihui Wu, Aohua Hunan Univ Coll Elect & Informat Engn Changsha Peoples R China State Grid Zhangjiajie Elect Power Supply Co Zhangjiajie Peoples R China

Power grid enterprise, as often as not, which is closely related to economic development is the core enterprises supporting regional economic development. In the field of the infrastructure planning of smart grid, the study of the excavation of development characteristics consolidates the theoretical basis for the formulation of distinguished source-network-loadstorage cost allocation mechanisms for power grids at different development levels [1][2]. The existing models and standards of appraisal which are incommensurate to the key technical characteristics of the backbone planning framework of smart grid with reliable grid connection of high percentage of renewable energy and power electronic equipment application, therefore, are in urgent need of being improved correspondingly [3][4]. This paper proposes a multiclosed-loop assessment mechanism for development level of smart grid with growth potential of power demand, growth potential of economic and growth potential of renewable energy penetration as error correction procedures, furthermore, and some opinions on the existing assessment methods are put forward. Based on the assessment results, some suggestions for distinguished infrastructure planning goals of smart grid at disparate development levels are presented [5].Considering the characteristics of "periodical" and "saturated" during the evolution process of power grid, this paper proposes a multi-closed-loop assessment mechanism for development level of smart grid. The key contributions of this opinion are twofold: 1) A multi-closed-loop assessment mechanism for development level of smart grid is formulated. Data analysis methods such as correlation analysis of the time difference and machine learning algorithm such as principal component analysis (PCA) is applied for an indicator which can comprehensively, objectively and accurately evaluate the current development level of the power grid is obtained. The generalized logistic growth curv

关键词： Supply Company Zhangjiajie smart grid renewable energy generalized logistic growth model dimension reduction algorithm cost sharing

来源：评论

学校读者我要写书评

暂无评论

Terahertz spectroscopy and machine learning algorithm for non-destructive evaluation of protein conformation

引用

OPTICAL AND QUANTUM ELECTRONICS 2020年第4期52卷 225-225页

作者： Cao, Can Zhang, Zhaohui Zhao, Xiaoyan Zhang, Tianyao Univ Sci & Technol Beijing Sch Automat & Elect Engn Beijing 100081 Peoples R China Beijing Engn Res Ctr Ind Spectrum Imaging Beijing 100081 Peoples R China

Given the condition that protein conformation and activity are highly susceptible to environment factors such as temperature and pH, evaluation of protein conformation and activity is urgently needed in many fields. For example, most protein drugs need a stable and proper environment during production, storage and transportation, and it's an enormous challenge to maintain protein activity throughout the whole process. Therefore, it's necessary to ensure the safety and effectiveness of protein drugs by monitoring their activity before use. In our study, we presented an improved method for non-destructive evaluation of protein conformation and biological activity by terahertz spectroscopy combined with t-SNE-XGBoost. Firstly, bovine serum albumin (BSA) samples heated to different temperature were measured with THz-TDS. The obtained results indicated that native-conformation BSA will undergo transient states in the process of temperature induced denaturation. However, for any single given sample, it's difficult to identify its conformation and activity directly by using the measured raw terahertz data. Therefore, we applied several different algorithms to the raw data for recognition of BSA samples with different conformation and activity induced by temperature. Finally, the models obtained by different algorithms were evaluated by calculating the root mean standard error of prediction (RMSEP) and the correlation coefficient of prediction (Rp). The THz-TDS plus t-SNE-XGBoost proved to be an effective non-destructive and label-free method for evaluation of protein conformation and activity. It can provide a new technique in many applications, such as pharmaceutical industry, clinical diagnosis and quality control.

关键词： Terahertz time-domain spectroscopy (THz-TDS) dimension reduction algorithm Machine learning algorithm Non-destructive evaluation Protein conformation

来源：评论

学校读者我要写书评

暂无评论

Visible-NIR spectral characterization and grade inversion modelling study of the Derni copper deposit

引用

INFRARED PHYSICS & TECHNOLOGY 2021年 115卷

作者： Mao, Yachun Huang, Jiaqi Cao, Wang Fu, Yanhua Zhao, Zhanguo Bao, Nisha Northeastern Univ Sch Resources & Civil Engn Shenyang 110819 Peoples R China Northeastern Univ Sch Architecture Shenyang 110819 Peoples R China China Gold Grp Beijing 100000 Peoples R China

At present, chemical analysis is the main method for copper ore grade verification. Because of its long test cycle and lag effect relative to mining and ore blending processes, this approach cannot effectively reduce losses in mineral resource mining. An effective way to solve this problem is to conduct research on in situ grade determination methods for ore bodies based on visible-near-infrared hyperspectral grade modelling techniques. However, the accuracy of the modelling is seriously reduced by the large redundancy of hyperspectral data and the self-limiting nature of some machine learning methods. Therefore, the accurate selection of dimensionality reduction algorithms and machine learning algorithms determines the accuracy of the inverse model, and the setting of parameters such as dimensionality reduction, number of nearest neighbour points and number of hidden layers in neural networks is crucial. In this paper, 190 ore samples were collected from the open pit of the Derni copper deposit in Dawu town, Guoluo Tibetan Autonomous Prefecture, Qinghai Province, China. First, the visible-near-infrared spectral data for the sample set were obtained with an SVC HR-1024 spectrometer. The copper grade of the sample set was verified by chemical analysis. Three algorithms-the Laplacian eigenmap (LE), local tangent space alignment (LTSA) and local linear embedding (LLE)-were used to reduce the dimensionality of the original hyperspectral data. The original data and the three hyperspectral datasets after dimensionality reduction were used as data sources. Several Derni copper grade inversion models based on two machine learning algorithms, the backpropagation neural network and radial basis function (RBF) neural network, were developed. Finally, the coefficient of determination (R2), mean absolute error (MAE) and root mean square error (RMSE) were used to evaluate the accuracy of each model. The results show that the accuracy of the model built with spectral data processe

关键词： Derni copper deposit Visible-near-infrared spectral data dimension reduction algorithm BP neural network RBF neural network

来源：评论

学校读者我要写书评

暂无评论

DOA Estimation algorithm for Wideband Polarization Sensitive Arrays

DOA Estimation Algorithm for Wideband Polarization Sensitive...

引用

IEEE Advanced Information Management, Communicates, Electronic and Automation Control Conference

作者： Bing Zhong Ning Wang CETC Key Laboratory of Intellisense Technology Nanjing China

Most existing array-based DOA estimation techniques only consider the effects of wideband features or polarization festures, and do not establish a unified signal model. In this paper, a general signal model of wideband polarization sensitive array is established by using wideband scalar array, narrowband polarization sensitive array signal model and DOA estimation technique. The ISM, CSM algorithm and the rank-loss dimension reduction MUSIC algorithm for polarization sensitive array are combined to obtain DOA estimation algorithm for wideband polarization sensitive arrays.

关键词： Polarization sensitive array Wideband signal DOA estimation dimension reduction algorithm

来源：评论

学校读者我要写书评

暂无评论

A Hybrid PCA-LDA Model for dimension reduction

A Hybrid PCA-LDA Model for Dimension Reduction

引用

International Joint Conference on Neural Networks (IJCNN)

作者： Zhao, Nan Mio, Washington Liu, Xiuwen Florida State Univ Dept Comp Sci Tallahassee FL 32306 USA Florida State Univ Dept Math Tallahassee FL 32306 USA

ISBN: (纸本)9781424496365

Several variants of Linear Discriminant Analysis (LDA) have been investigated to address the vanishing of the within-class scatter under projection to a low-dimensional subspace in LDA. However, some of these proposals are ad hoc and some others do not address the problem of generalization to new data. Meanwhile, even though LDA is preferred in many application of dimension reduction, it does not always outperform Principal Component Analysis (PCA). In order to optimize discrimination performance in a more generative way, a hybrid dimension reduction model combining PCA and LDA is proposed in this paper. We also present a dimension reduction algorithm correspondingly and illustrate the method with several experiments. Our results have shown that the hybrid model outperform PCA, LDA and the combination of them in two separate stages.

关键词： Computational modeling Cost function Data models Principal component analysis Training Training data Vectors data analysis dimension reduction algorithm discrimination performance face recognition hybrid PCA-LDA model hybrid dimension reduction model learning (artificial intelligence) linear discriminant analysis low-dimensional subspace principal component analysis principal component analysis within-class scatter under projection

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：