检索结果-内蒙古大学图书馆

IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM)

作者： Osho, Abiola Waters, Caden Amariucai, George Kansas State Univ Dept Comp Sci Manhattan KS 66506 USA

ISBN: (纸本)9781728110561

With the increasing use of online social networks as a source of news and information, the propensity for a rumor to disseminate widely and quickly poses a great concern, especially in disaster situations where users do not have enough time to fact-check posts before making the informed decision to react to a post that appears to be credible. At the same time, we know that misinformation is easily detectable by a certain few, very skeptical, or very informed users. In this study, we demonstrate how blending artificial intelligence and human skills can create a new paradigm for credibility prediction. The crowdsourcing part of the detection mechanism is implemented implicitly, by simply observing the natural interaction between users encountering the messages. Specifically, we explore the spread of information on Twitter at the microscopic (user-to-user propagation) level and propose a model that predicts if a message is True or False by observing the latent attributes of the message, along with those of the users interacting with it, and their reactions to the message. We demonstrate the application of this model to the detection of misinformation and rank the relevant message and user features that are most critical in influencing the spread of rumor over the network. Our experiments using real-world data show that the proposed model achieves over 90% accuracy in predicting the credibility of posts on Twitter, a significant boost over stateof-the-art models.

关键词： Crowdsourcing Social Networks Bayesian Learning classification and regression Misinformation Dimensionality Reduction/Feature Selection

来源：评论

学校读者我要写书评

暂无评论

Can Shallow Neural Networks Beat the Curse of Dimensionality? A Mean Field Training Perspective

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2020年第2期1卷 121-129页

作者： Wojtowytsch, Stephan Weinan, E. The Program of Applied and Computational Mathematics Princeton University Princeton 08544 NJ United States The Department of Mathematics The Program of Applied and Computational Mathematics Princeton University Princeton 08544 NJ United States

We prove that the gradient descent training of a two-layer neural network on empirical or population risk may not decrease population risk at an order faster than t−4/(d−2) under mean field scaling. The loss functiona... 详细信息

We prove that the gradient descent training of a two-layer neural network on empirical or population risk may not decrease population risk at an order faster than t⁻⁴/(^d−2) under mean field scaling. The loss functional is mean squared error with a Lipschitz-continuous target function and data distributed uniformly on the d-dimensional unit cube. Thus gradient descent training for fitting reasonably smooth, but truly high-dimensional data may be subject to the curse of dimensionality. We present numerical evidence that gradient descent training with general Lipschitz target functions becomes slower and slower as the dimension increases, but converges at approximately the same rate in all dimensions when the target function lies in the natural function space for two-layer ReLU networks. Impact Statement–Artificial neural networks perform well in many real life applications, but may suffer from the curse of dimensionality on certain problems. We provide theoretical and numerical evidence that this may be related to whether a target function lies in the hypothesis class described by infinitely wide networks. The training dynamics are considered in the fully non-linear regime and not reduced to neural tangent kernels. We believe that it will be essential to study these hypothesis classes in detail to choose an appropriate machine learning models for a given problem. The goal of the article is to illustrate this in a mathematically sound and numerically convincing fashion. © 2021 IEEE.

关键词： Artificial neural networks classification and regression supervised learning

来源：评论

学校读者我要写书评

暂无评论

Sparse and heuristic support vector machine for binary classifier and regressor fusion

引用

INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS 2019年第12期10卷 3667-3686页

作者： Huang, Jinhong Yu, Zhu Liang Gu, Zhenghui Zhang, Jun Cen, Ling South China Univ Technol Coll Automat Sci & Engn Guangzhou Guangdong Peoples R China Guangdong Univ Technol Coll Informat Engn Guangzhou Guangdong Peoples R China

Single-hidden layer feedforward networks (SLFNs) are always viewed as classical methods for binary classification and regression. There are several variant types of SLFNs, such as support vector machines (SVM), extreme learning machines (ELM), etc. It is an open problem for SLFNs to obtain a powerful feature mapper with a simple network structure. In this paper, we propose a framework called sparse and heuristic SVM (SH-SVM) to fuse different SLFNs from the aspect of feature mapping to obtain powerful feature mapping capability and improve the generalization performance. By fusing different SLFNs, the SH-SVM benefits from the learning capabilities of each model. As an example, the fusion of SVM and ELM is studied in detail. Then with the sparse representation method, a compact SLFN is obtained and the most powerful hidden nodes are selected. Furthermore, an efficient method for solving the sparse representation problem in SH-SVM is proposed. Experiments on 25 data sets with eight methods show that SH-SVM has satisfactory results with a compact network structure.

关键词： Single hidden layer feedforward networks Extreme learning machine Support vectors sparse and heuristic classification and regression

来源：评论

学校读者我要写书评

暂无评论

Prediction of Gender and Age from Inertial Sensor-based Gait Dataset 8

Prediction of Gender and Age from Inertial Sensor-based Gait...

引用

Joint 8th International Conference on Informatics, Electronics and Vision (ICIEV) / 3rd International Conference on Imaging, Vision and Pattern Recognition (icIVPR) / International Conference on Activity and Behavior Computing (ABC)

作者： Khabir, Kanij Mehtanin Siraj, Md Sadman Ahmed, Masud Ahmed, Mosabber Uddin Univ Dhaka Dept Elect & Elect Engn Dhaka Bangladesh

ISBN: (纸本)9781728107882

The technology pertaining to gait generation, recognition and analysis is getting more sophisticated each day. The global demand for a gait-based dataset and its subsequent recognition, leading to extraction of valuable information is now higher than ever. However, inertial sensor-based gait dataset is a comparatively new addition to the field of gait analysis. Consequently, most of the research works incorporating machine learning algorithms into gait dataset are image based. In addition to that, most of the gait-based datasets have been analyzed for gait recognition. There remains very little research work on personal authentication from inertial sensor-based gait dataset. Personal authentication is of several types out of which, predicting gender and age is quite challenging. In this paper, we have tried to face these challenges and have manifested the process of predicting gender and age from the inertial sensor-based gait dataset which is a part of the vast Osaka University-ISIR Gait Database. Finally, we have found that the models, Support Vector Machine shows the highest accuracy for the problem of classifying gender and Decision Trees shows the highest variance score (R-2 value) for the problem of predicting age.

关键词： gait inertial sensor supervised machine learning gender and age estimation classification and regression

来源：评论

学校读者我要写书评

暂无评论

Global Terrorism Predictive-Analysis

Global Terrorism Predictive-Analysis

引用

International Conference on Computer, Communication and Computational Sciences (IC4S)

作者： Chaurasia, Sandeep Warikoo, Vinayak Khan, Shanu Manipal Univ Jaipur Jaipur Rajasthan India Indian Inst Technol Roorkee Roorkee Uttar Pradesh India

ISBN: (纸本)9789811368615;9789811368608

In the recent view of increasing number and lethality of terrorist attacks, it has become important for us to recognize a strategic vision to help prepare and prevent such events from happening. This paper includes descriptive and predictive analyses of Global Terrorism Database which reveal vital information about the trends of such events and help identify the perpetrators of any such future terrorist activities. The descriptive phase covers elucidation of the dataset to identify useful features for forecasting and predictive phase involves data manipulation and compares the performance of various multi-class classification and regression algorithms like decision trees, random forest, etc., on the dataset. Python with Scikit-learn library was used for the experimentation purpose.

关键词： Global terrorism database Descriptive analysis Predictive analysis classification and regression

来源：评论

学校读者我要写书评

暂无评论

Deep Learning for real-time gravitational wave detection and parameter estimation: Results with Advanced LIGO data

引用

PHYSICS LETTERS B 2018年 778卷 64-70页

作者： George, Daniel Huerta, E. A. Univ Illinois Dept Astron Urbana IL 61801 USA Univ Illinois NCSA Urbana IL 61801 USA

The recent Nobel-prize-winning detections of gravitational waves from merging black holes and the subsequent detection of the collision of two neutron stars in coincidence with electromagnetic observations have inaugurated a new era of multimessenger astrophysics. To enhance the scope of this emergent field of science, we pioneered the use of deep learning with convolutional neural networks, that take time-series inputs, for rapid detection and characterization of gravitational wave signals. This approach, Deep Filtering, was initially demonstrated using simulated LIGO noise. In this article, we present the extension of Deep Filteringusing real data from LIGO, for both detection and parameter estimation of gravitational waves from binary black hole mergers using continuous data streams from multiple LIGO detectors. We demonstrate for the first time that machine learning can detect and estimate the true parameters of real events observed by LIGO. Our results show that Deep Filtering achieves similar sensitivities and lower errors compared to matched- filtering while being far more computationally efficient and more resilient to glitches, allowing real-time processing of weak time-series signals in non-stationary non-Gaussian noise with minimal resources, and also enables the detection of new classes of gravitational wave sources that may go unnoticed with existing detection algorithms. This unified framework for data analysis is ideally suited to enable coincident detection campaigns of gravitational waves and their multimessenger counterparts in real-time. (C) 2018 The Authors. Published by Elsevier B.V.

关键词： Deep Learning Convolutional neural networks Gravitational waves LIGO Time-series signal processing classification and regression

来源：评论

学校读者我要写书评

暂无评论

Remote Sensing Based Modeling of Dengue Outbreak with regression and Binning classification 2

Remote Sensing Based Modeling of Dengue Outbreak with Regres...

引用

2nd IEEE International Conference on Computer and Communications (ICCC)

作者： Kerdprasop, Nittaya Kerdprasop, Kittisak Suranaree Univ Technol Data Engn Res Unit Sch Comp Engn Nakhon Ratchasima Thailand Suranaree Univ Technol Knowledge Engn Res Unit Sch Comp Engn Nakhon Ratchasima Thailand

ISBN: (纸本)9781467390262

We present in this paper the application of data mining techniques using remote sensing data to help modeling the prominent characteristics of dengue and severe dengue outbreak in the northeast of Thailand. Dengue is a lethal mosquito-borne viral disease that has long been a serious public health problem in many tropical and some sub-tropical countries. We investigate the rainfall and other remote sensing variables to correlate their contributions to the 2006-2015 dengue outbreaks in Nakhon Ratchasima province through the regression analysis. Classical and severe dengue cases during the past ten years are then discretized with the two binning methods (equal-width 4 bins and standard deviation based 3 bins) and used as the target for model building. Models are built as the classification and regression tree using CART and C5 algorithms. With the cross validation evaluation method, the models built from the equal-width 4 bins show correct classification performance in the range of 78-86%, whereas the models built from the standard deviation based 3 bins yield the 91-95% accuracy.

关键词： dengue outbreak modeling classification and regression binning method remote sensing data

来源：评论

学校读者我要写书评

暂无评论

Associations of Coexisting Conditions with Healthcare Spending for Children with Cerebral Palsy

引用

JOURNAL OF PEDIATRICS 2018年 200卷 111-+页

作者： Berry, Jay G. Glader, Laurie Stevenson, Richard D. Hasan, Fareesa Crofton, Charis Hussain, Kinza Hall, Matt Boston Childrens Hosp Dept Med Div Gen Pediat Boston MA USA Harvard Med Sch Boston MA USA Univ Virginia Sch Med Div Dev Pediat Charlottesville VA 22908 USA Boston Univ Sch Med Div Grad Med Sci Boston MA 02118 USA Boston Univ Sch Publ Hlth Boston MA 02215 USA Massachusetts Coll Pharm & Hlth Sci Boston MA USA Childrens Hosp Assoc Lenexa KS USA

Objective To determine which coexisting conditions have the strongest associations with healthcare use and spending among children with cerebral palsy (CP). Study design Retrospective analysis of 16 695 children ages 0-18 years with CP - identified with International classification of Diseases, Ninth Revision, Clinical Modification codes - using Medicaid from January 1, 2013 to December 31, 2013 from 10 states in the Truven MarketScan Medicaid Database. Using generalized linear models, we assessed which coexisting conditions (including medical technology) identified with Agency for Healthcare Research and Quality's Chronic Condition Indicators had the strongest associations with total healthcare spending across the healthcare continuum. Results Median per-patient annual Medicaid spending for children with CP was $12 299 (IQR $4826-$35 582). Most spending went to specialty (33.1%) and hospital (26.7%) care. The children had a median 6 (IQR 4-10) coexisting conditions;epilepsy was the most common (38.1%). Children with epilepsy accounted for 59.6% ($364 million) of all CP spending. In multivariable analysis, the coexisting conditions most strongly associated with increased spending were tracheostomy (median additional cost per patient = $56 567 [95%CI $51 386-61 748]) and enterostomy (median additional cost per patient = $25 707 [95%CI $23 753-27 660]). Conclusions Highly prevalent in children with CP using Medicaid, coexisting conditions correlate strongly with healthcare spending. Tracheostomy and enterostomy, which indicate significant functional impairments in breathing and digestion, are associated with the highest spending. Families, providers, payers. and legislators may leverage these findings when designing policies positioned to enable the best health and care for children with cerebral palsy.

关键词： Medicaid disability cost utilization AHRQ Agency for Healthcare Research and Quality CART classification and regression CCI Chronic condition indicator CP Cerebral palsy DME Durable medical equipment GMFCS Gross Motor Function classification System ICD-9-CM International classification of Diseases, Ninth Revision, Clinical Modification

来源：评论

学校读者我要写书评

暂无评论

Factors associated with success of telaprevir-and boceprevir-based triple therapy for hepatitis C virus infection

引用

World Journal of Hepatology 2017年第11期9卷 551-561页

作者： Kian Bichoupan Neeta Tandon Valerie Martel-Laferriere Neal M Patel David Sachs Michel Ng Emily A Schonfeld Alexis Pappas James Crismale Alicia Stivala Viktoriya Khaitova Donald Gardenier Michael Linderman William Olson Ponni V Perumalswami Thomas D Schiano Joseph A Odin Lawrence U Liu Douglas T Dieterich Andrea D Branch Division of Liver Diseases Icahn School of Medicine at Mount Sinai Icahn Medical Instititute Janssen Pharmaceuticals Com-panies Division of Janssen Scientific Affairs Inc. Genetics and Genomics Icahn School of Medicine at Mount Sinai

To evaluate new therapies for hepatitis C virus (HCV), data about real-world outcomes are needed. METHODSOutcomes of 223 patients with genotype 1 HCV who started telaprevir- or boceprevir-based triple therapy (May 2011-March 2012) at the Mount Sinai Medical Center were analyzed. Human immunodeficiency virus-positive patients and patients who received a liver transplant were excluded. Factors associated with sustained virological response (SVR24) and relapse were analyzed by univariable and multivariable logistic regression as well as classification and regression trees. Fast virological response (FVR) was defined as undetectable HCV RNA at week-4 (telaprevir) or week-8 (boceprevir). RESULTSThe median age was 57 years, 18% were black, 44% had advanced fibrosis/cirrhosis (FIB-4 ≥ 3.25). Only 42% (94/223) of patients achieved SVR24 on an intention-to-treat basis. In a model that included platelets, SVR24 was associated with white race [odds ratio (OR) = 5.92, 95% confidence interval (CI): 2.34-14.96], HCV sub-genotype 1b (OR = 2.81, 95%CI: 1.45-5.44), platelet count (OR = 1.10, per x 10⁴ cells/μL, 95%CI: 1.05-1.16), and IL28B CC genotype (OR = 3.54, 95%CI: 1.19-10.53). Platelet counts > 135 x 10³/μL were the strongest predictor of SVR by classification and regression tree. Relapse occurred in 25% (27/104) of patients with an end-of-treatment response and was associated with non-FVR (OR = 4.77, 95%CI: 1.68-13.56), HCV sub-genotype 1a (OR = 5.20; 95%CI: 1.40-18.97), and FIB-4 ≥ 3.25 (OR = 2.77; 95%CI: 1.07-7.22). CONCLUSIONThe SVR rate was 42% with telaprevir- or boceprevir-based triple therapy in real-world practice. Low platelets and advanced fibrosis were associated with treatment failure and relapse.

关键词： Sustained virologic response Hepatitis C virus Relapse Telaprevir Boceprevir Triple-therapy classification and regression Adverse event Real-world

来源：评论

学校读者我要写书评

暂无评论

Predicting Higher Secondary Results by Data Mining Algorithms with VBR: A feature Reduction Method 4

Predicting Higher Secondary Results by Data Mining Algorithm...

引用

4th International Conference on Advances in Electrical Engineering (ICAEE)

作者： Shahadat, Nazmul Rahman, Musfiqur Ahmed, Sakil Rahman, Bushra Prime Univ Dhaka Bangladesh City Univ Dhaka Bangladesh

ISBN: (纸本)9781538608692

The Higher Secondary Certificate (HSC) Examination in Bangladesh was very crucial for students' future, as well as job lives. As the performance of this examination was influenced by some factors, the main scope of this research was to identify these influencing factors with highest possible precision. In this research, an experimental methodology was adopted to generate a dataset which was constructed from students who passed HSC. Several Data mining methods were used to analyze data, extracting models, and develop knowledge to make decision and also used some feature selection methods to remove unnecessary and redundant features which were helped to improve the performance of the classifiers. Here applied several existent feature selection algorithms and a proposed approach to find the relevant features. The results of the empirical analysis were obtained by removing these irrelevant features with some bayes based, function based, lazy based, rule base and tree base classifiers. And finally the results showed, feature reduction had some effect on performance and for overall LMT classifier with our selected features performed better than others.

关键词： Higher Secondary Student performance Higher Secondary Education feature selection Vital features Data Mining classification and regression LMT

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：