检索结果-内蒙古大学图书馆

Heart Failure Prediction Based on bootstrap sampling and Weighted Fusion LightGBM Model

APPLIED SCIENCES-BASEL 2025年第8期15卷 4360-4360页

作者： Wang, Yuanni Cao, Hong China Univ Geosci Sch Comp Sci Wuhan 430078 Peoples R China China Univ Geosci Hubei Key Lab Intelligent Geoinformat Proc Wuhan 430078 Peoples R China

Heart disease is a serious threat to human health. Accurate prediction is very important for disease prevention and treatment. The purpose of this study is to establish a more suitable prediction model of heart disease. Based on LightGBM, we have deeply integrated bootstrap sampling and weighting technology. We repeatedly use multiple parameters of LightGBM to perform bootstrap sampling on the original training set. Through this process, we not only obtain training sub-models for various training subsets but also mine rich data features. In the process of cross-validation, the weight coefficient for each sub-model is carefully determined by comprehensively evaluating multiple key performance indicators, including accuracy, precision, recall, and the F1 score. This approach effectively highlights the role of high-quality sub-models. In the test stage, each sub-model is weighted according to the weight corresponding to its specific parameter combination, and finally, an accurate prediction result is obtained. Compared with the traditional prediction model, the model shows better comprehensive performance in terms of various performance metrics such as accuracy, precision, recall, and F1 score, and also performs better in the paired t-test. Moreover, compared with the baseline model, the phenomenon of overfitting is obviously reduced. Although the model has not been verified by external data sets, it has, to a certain degree, significantly boosted its predictive ability, universality, and stability. Moreover, it has provided a feasible scheme for heart disease prediction, which is expected to play a crucial role in clinical auxiliary diagnosis and disease management. The research shows that this model has obvious advantages in heart disease prediction and can effectively enhance the accuracy and reliability of prediction.

关键词： heart disease prediction LightGBM bootstrap sampling weighted fusion

来源：评论

学校读者我要写书评

暂无评论

bootstrap sampling based Data Cleaning and Maximum Entropy SVMs for large Datasets

Bootstrap Sampling based Data Cleaning and Maximum Entropy S...

引用

IEEE 24th International Conference on Tools with Artificial Intelligence (ICTAI)

作者： Wang, Senzhang Li, Zhoujun Zhang, Xiaoming Beihang Univ State Key Lab Software Dev Environm Beijing Key Lab Network Technol Beijing Peoples R China

ISBN: (纸本)9780769549156;9781479902279

Support Vector Machines (SVMs) is a popular machine learning algorithm based on Statistical Learning Theory (SLT). However, traditional solutions suffer from O(n(2)) time complexity. In this paper, a novel two-stage informative pattern abstraction algorithm is proposed. The first stage of the algorithm is data cleaning based on bootstrap sampling. A bundle of weak SVM classifiers are trained based on the sampled small datasets. Training data correctly classified by all the weak classifiers are cleaned. In the second stage, to further improve performance of final classifier and reduce training time, two novel informative pattern extraction algorithms based on entropy maximization SVMs are proposed. Empirical studies show our approach is effective in reducing size of training datasets and the computational cost, outperforming the state-of-the-art SVM training algorithms PEGASOS, RSVM and LIBLINEAR SVM with comparable classification accuracy.

关键词： SVMs bootstrap sampling entropy maximization

来源：评论

学校读者我要写书评

暂无评论

Non-parametric and unsupervised Bayesian classification with bootstrap sampling

引用

IMAGE AND VISION COMPUTING 2004年第1期22卷 1-8页

作者： Zribi, M Univ Littoral Cote Opale Maison Rech Blaise Pascal LASL UPRES 2600 F-62228 Calais France

In this paper, we propose a non-parametric and unsupervised Bayesian classification based on the principle of bootstrap sampling (BS) which reduces the dependence effect of pixels in real images, and reduces the classification time. Given an original image, we randomly select a small representative set of pixels. Then, a Non-parametric Expectation-Maximization (NEM) algorithm is used for image segmentation. The non-parametric aspect comes from the use of the orthogonal probability density function (pdf) estimation, which is reduced to the estimation of the first Fourier coefficients (FC's) of the pdf with respect to a given orthogonal basis. The results we obtain show that the BS method gives better results than the classical one, both in the quality of the segmented image and the computing time. (C) 2003 Elsevier B.V. All rights reserved.

关键词： non-parametric and unsupervised Bayesian classification bootstrap sampling non-parametric expectation-maximization orthogonal probability density functions

来源：评论

学校读者我要写书评

暂无评论

A high spatial resolution residential energy model based on American Time Use Survey data and the bootstrap sampling method

引用

ENERGY AND BUILDINGS 2011年第12期43卷 3528-3538页

作者： Chiou, Yun-Shang Carley, Kathleen M. Davidson, Cliff I. Johnson, Michael P. Natl Taiwan Univ Sci & Technol Dept Architecture Taipei 106 Taiwan Carnegie Mellon Univ Sch Architecture Pittsburgh PA 15213 USA Carnegie Mellon Univ Ctr Computat Anal Social & Org Syst Pittsburgh PA 15213 USA Carnegie Mellon Univ Inst Software Res Sch Comp Sci Pittsburgh PA 15213 USA Syracuse Univ Ctr Excellence Environm & Energy Syst Syracuse NY 13244 USA Syracuse Univ Dept Civil & Environm Engn Syracuse NY 13244 USA Univ Massachusetts McCormack Grad Sch Policy Studies Dept Publ Policy & Publ Affairs Boston MA 02125 USA

A high spatial resolution model of energy use in residential buildings was developed based on time use data. The development of this model was guided by a theoretical framework that explains the nature of a dwelling's physical characteristics, the nature of its occupants' energy use behaviors and the ways in which the dwelling and its occupants interact to determine the energy use of a household. Energy Plus was used as the model platform. In the model, the occupants' domestic activity pattern was extracted from American Time Use Survey (ATUS) data via bootstrap sampling. The dwelling's physical characteristics were based on real world scenarios. Virtual experiments with 3- to 5-occupant household compositions were conducted to examine the model properties. Simulation results show that (1) bell-shaped distributions were present in annual heating load demands for all household compositions, (2) the load demands for different batches of samples of the same household composition demonstrated a narrow range of variations, and (3) the simulated hourly appliance and lighting load profiles were in agreement with those generated from field-metered data at both whole-house and sub-house levels. These results indicate the model's overall robustness and verify its ability to simulate realistic residential energy use load profiles. (C) 2011 Elsevier B.V. All rights reserved.

关键词： American Time Use Survey (ATUS) bootstrap sampling Residential energy model Human behavior Sub-house spatial resolution

来源：评论

学校读者我要写书评

暂无评论

Non-parametric and region-based image fusion with bootstrap sampling

引用

INFORMATION FUSION 2010年第2期11卷 85-94页

作者： Zribi, Mourad Univ Littoral Cote dOpale LASL UPRES 2600 F-62228 Calais France

image fusion refers to the techniques that integrate complementary information from multiple image sensors' data in a way that makes the new images more suitable for human Visual perception and reduces computation processing tasks. In this paper, we propose a non-parametric and region-based image fusion based oil the bootstrap sampling (BS) principle, which reduces the dependence effect of pixels in real images and minimizes the fusion time. Given an original image, we randomly select a small representative set of pixels In the statistical image formation model, image sensors are described as the true scene corrupted by additive non-Gaussian distortion Then, a Non-parametric Expectation-Maximization (NEM) algorithm would be used to estimate both the model parameters and the fused image. The non-parametric aspect comes from the use of the orthogonal series' estimator. Obtained results show that the BS method gives better results than the classical one, both for fused image quality as well as for computation time. (C) 2008 Elsevier B.V All rights reserved.

关键词： Region-based image fusion bootstrap sampling Non-parametric Expectation-Maximization Orthogonal series estimator

来源：评论

学校读者我要写书评

暂无评论

Performing competitive analysis in QFD studies using state multipole moments and bootstrap sampling

引用

QUALITY ENGINEERING 2017年第2期29卷 311-321页

作者： Iqbal, Zafar Grigg, Nigel P. Govindaraju, K. Massey Univ Sch Engn & Adv Technol Palmerston North New Zealand

QFD often makes use of Likert scales to convert opinions into data for analysis. Likert scales are prone to bias in relation to their labeling, interpretation, and distributional assumptions. Analysis is often based on means, without regard to variation or distribution shape, making inferential statistical procedures unreliable. We present a method for extracting increased information from Likert data using State Multipole Moments, and use bootstrap sampling to enable the moments to be statistically compared. We demonstrate application using a published QFD competitor analysis, showing how this method can enable practitioners to perform statistical comparisons resulting in improved decision making.

关键词： quality function deployment (qfd) bootstrap sampling competitive analysis state multipole moments likert scale data

来源：评论

学校读者我要写书评

暂无评论

Determining the Best Clustering Number of K-means Based on bootstrap sampling 2

Determining the Best Clustering Number of K-means Based on B...

引用

2nd International Conference on Data Science and Business Analytics (ICDSBA)

作者： Yu, Lianmin Zhou, Changyin Shandong Univ Sci & Technol Coll Math & Syst Sci Qingdao Shandong Peoples R China

ISBN: (纸本)9781538684313

In this paper, the selection of clustering number in K-means clustering algorithm is studied, based on the bootstrap sampling, a new method is proposed to determine the best clustering number based on the between the actual value of total within-cluster sum of squares and its estimated interval. By UCI Machine Learning Repository and randomly generated artificial simulated test data sets, the experimental results show that using clustering has obvious improvement, the method can overcome falls into local optimum caused by unreasonable clustering number to select. K-means algorithm.

关键词： K-means clustering algorithm the best clustering number within-cluster sum of squares bootstrap sampling

来源：评论

学校读者我要写书评

暂无评论

Stochastic Conceptual Cost Estimating of Highway Projects to Communicate Uncertainty Using bootstrap sampling

引用

ASCE-ASME JOURNAL OF RISK AND UNCERTAINTY IN ENGINEERING SYSTEMS PART A-CIVIL ENGINEERING 2017年第3期3卷

作者： Gardner, Brendon J. Gransberg, Douglas D. Rueda, Jorge A. Iowa State Univ Dept Civil Construct & Environm Engn Ames IA 50011 USA

Conceptual cost estimating is typically completed early in the project lifecycle when little design work has been completed. Because little information is known at this early stage, the estimate usually deviates substantially from the actual construction cost. When expressed as a deterministic value, an estimate often leads to a false inference of accuracy by those not familiar with the vagaries of conceptual cost estimating, making it difficult for an agency to explain cost growth. A stochastic conceptual estimate allows an agency to produce a probability distribution of the likely construction costs and address the level of confidence in a given estimate. Named probability distributions are readily available for developing a stochastic estimate in a great deal of commercial software. However, instead of fitting available distributions, this research generates an empirical distribution to express a cost estimate range. Creating empirical distributions eliminates assumptions required for selecting named distributions. The stochastic data-driven model developed in this paper combines artificial neural networks and bootstrap sampling in a case study of 189 highway projects to train and test the estimating model. (C) 2016 American Society of Civil Engineers.

关键词： Cost estimating Risk Range estimating Confidence level Stochastic bootstrap sampling Artificial neural network Cost transparency

来源：评论

学校读者我要写书评

暂无评论

Comparison of Iterative-Two Stage and bootstrap sampling Approaches in the Development of a Population Pharmacokinetic Model

引用

IFAC Proceedings Volumes 1997年第2期30卷 141-145页

作者： Rajeshwari Sridhara Mario A. Eisenberger Leonard M. Reyno Duncan I. Jodrell Victoria J. Sinibaldi Eleanor G. Zubowski Merrill J. Egorin Baltimore MD 21201 Baltimore MD 21287 Hamilton Ontario L8V5C2 Canada Edinburgh EH42XLI U.K.

The kinetics of an antineoplastic drug, suramin, described by a three-compartment, open, linear model is studied in this report. A population pharmacokinetic model, describing the pharmacokinetics of suramin, was developed using the iterative-two stage approach, with maximum a posteriori-Bayesian priors. It was further validated using data from a subsequent study (validation data set). The population pharmacokinetic parameters were also estimated by generating bootstrap samples from the validation data set. When the prior distribution was close to the population distribution, there was no significant difference between the estimates obtained using the iterative-two stage approach and the bootstrap approach. However, when estimating population pharmacokinetic parameters, the iterative-two stage approach was less sensitive to prior distribution specification

关键词： Population pharmacokinetics suramin iterative-two stage bootstrap sampling

来源：评论

学校读者我要写书评

暂无评论

Assessing genomic prediction accuracy for Holstein sires using bootstrap aggregation sampling and leave-one-out cross validation

引用

JOURNAL OF DAIRY SCIENCE 2017年第1期100卷 453-464页

作者： Mikshowsky, Ashley A. Gianola, Daniel Weigel, Kent A. Univ Wisconsin Dept Dairy Sci Madison WI 53706 USA Univ Wisconsin Dept Anim Sci Madison WI 53706 USA Univ Wisconsin Dept Biostat & Med Informat Madison WI 53706 USA

Since the introduction of genome-enabled prediction for dairy cattle in 2009, genomic selection has markedly changed many aspects of the dairy genetics industry and enhanced the rate of response to selection for most economically important traits. Young dairy bulls are genotyped to obtain their genomic predicted transmitting ability (GPTA) and reliability (REL) values. These GPTA are a main factor in most purchasing, marketing, and culling decisions until bulls reach 5 yr of age and their milk-recorded offspring become available. At that time, daughter yield deviations (DYD) can be compared with the GPTA computed several years earlier. For most bulls, the DYD align well with the initial predictions. However, for some bulls, the difference between DYD and corresponding GPTA is quite large, and published REL are of limited value in identifying such bulls. A method of bootstrap aggregation sampling (bagging) using genomic BLUP (GBLUP) was applied to predict the GPTA of 2,963, 2,963, and 2,803 young Holstein bulls for protein yield, somatic cell score, and daughter pregnancy rate (DPR), respectively. For each trait, 50 bootstrap samples from a reference population comprising 2011 DYD of 8,610, 8,405, and 7,945 older Holstein bulls were used. Leave-one-out cross validation was also performed to assess prediction accuracy when removing specific bulls from the reference population. The main objectives of this study were (1) to assess the extent to which current REL values and alternative measures of variability, such as the bootstrap standard deviation (SD) of predictions, could detect bulls whose daughter performance deviates significantly from early genomic predictions, and (2) to identify factors associated with the reference population that inform about inaccurate genomic predictions. The SD of bootstrap predictions was a mildly useful metric for identifying bulls whose future daughter performance may deviate significantly from early GPTA for protein and DPR. Leave-one

关键词： genomic prediction leave-one-out cross validation bootstrap sampling dairy cattle

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：