检索结果-内蒙古大学图书馆

An application of machine learning with feature selection to improve diagnosis and classification of neurodegenerative disorders

引用

BMC BIOINFORMATICS 2019年第1期20卷 1-12页

作者： Diaz Alvarez, Josefa Matias-Guiu, Jordi A. Nieves Cabrera-Martin, Maria Risco-Martin, Jose L. Ayala, Jose L. Univ Extremadura Dept Comp Architecture & Commun Merida Badajoz Spain Univ Complutense Dept Neurol Hosp Clin San Carlos San Carlos Res Hlth Inst IdISSC Madrid Spain Univ Complutense Dept Comp Architecture & Automat Madrid Spain

Background: The analysis of health and medical data is crucial for improving the diagnosis precision, treatments and prevention. In this field, machine learning techniques play a key role. However, the amount of health data acquired from digital machines has high dimensionality and not all data acquired from digital machines are relevant for a particular disease. Primary Progressive Aphasia (PPA) is a neurodegenerative syndrome including several specific diseases, and it is a good model to implement machine learning analyses. In this work, we applied five feature selection algorithms to identify the set of relevant features from 18F-fluorodeoxyglucose positron emission tomography images of the main areas affected by PPA from patient records. On the other hand, we carried out classification and clustering algorithms before and after the feature selection process to contrast both results with those obtained in a previous work. We aimed to find the best classifier and the more relevant features from the WEKA tool to propose further a framework for automatic help on diagnosis. Dataset contains data from 150 FDG-PET imaging studies of 91 patients with a clinic prognosis of PPA, which were examined twice, and 28 controls. Our method comprises six different stages: (i) feature extraction, (ii) expertise knowledge supervision (iii) classification process, (iv) comparing classification results for feature selection, (v) clustering process after feature selection, and (vi) comparing clustering results with those obtained in a previous work. Results: Experimental tests confirmed clustering results from a previous work. Although classification results for some algorithms are not decisive for reducing features precisely, Principal Components Analisys (PCA) results exhibited similar or even better performances when compared to those obtained with all features. Conclusions: Although reducing the dimensionality does not means a general improvement, the set of features is almost hal

关键词： Machine learning primary progressive aphasia Supervised algorithm unsupervised algorithm Clustering Analysis

来源：评论

学校读者我要写书评

暂无评论

A self-supervised strategy for fully automatic segmentation of renal dynamic contrast-enhanced magnetic resonance images

引用

MEDICAL PHYSICS 2019年第10期46卷 4417-4430页

作者： Huang, Wenjian Li, Hao Wang, Rui Zhang, Xiaodong Wang, Xiaoying Zhang, Jue Peking Univ Acad Adv Interdisciplinary Studies Beijing Peoples R China Peking Univ Hosp 1 Dept Radiol Beijing Peoples R China Peking Univ Coll Engn Beijing Peoples R China

Purpose An automated accurate segmentation for dynamic contrast-enhanced magnetic resonance (DCE-MR) image sequences is essential for quantification of renal function. A self-supervised strategy is proposed for fully automatic segmentation of the renal DCE-MR images without using manually labeled data. Methods The proposed strategy employed both temporal and spatial information of the DCE-MR image sequences. First, the kidney area, the seed regions of the cortex, the medulla, and the pelvis were automatically detected in the spatial domain. Subsequently, all the pixels in the kidney were automatically labeled as the cortex, the medulla, or the pelvis based on their time-intensity signal and spatial position using a supervised classifier. The feasibility of the proposed strategy was verified on a dataset of renal DCE-MR images of 14 subjects without history of kidney diseases. Furthermore, the self-supervised strategy and the commonly used traditional unsupervised method were quantitatively compared with a reference manual segmentation by an experienced radiologist, using similarity indexes. Results The average Dice coefficient (ADC) for the segmentations of the proposed self-supervised method is 0.92 using a ransom walker model as the classifier or 0.86 using a K-nearest neighbor model as the classifier. The ADC of the Kmeans-based unsupervised methods with three and six clusters were 0.65 and 0.79, respectively. The Dice coefficients of the self-supervised method were remarkably higher than that of the unsupervised method (one-tailed paired-sample t-test, P-values <10(-3)). Conclusions The results indicate that the proposed self-supervised approach yields a satisfactory similarity with the reference manual segmentation. Compared with the traditional unsupervised clustering method, the new strategy does not require manual intervention during the segmentation process and achieves better results for the segmentation of renal DCE-MR images.

关键词： automatic kidney segmentation DCE-MRI random walker self-supervised algorithm unsupervised algorithm

来源：评论

学校读者我要写书评

暂无评论

A Framework for Hoax News Detection and Analyzer used Rule-based Methods

引用

INTERNATIONAL JOURNAL OF ADVANCED COMPUTER SCIENCE AND APPLICATIONS 2019年第10期10卷 402-408页

作者： Yuliani, S. Y. Bin Abdollah, Mohd Faizal Sahib, Shahrin Wijaya, Yunus Supriadi Univ Tekn Malaysia Melaka Fac Informat Commun Technol Informat Secur & Networking Res Grp InFORSNET Melaka Malaysia Widyatma Univ Bandung Indonesia Univ Tekn Malaysia Melaka Fac Informat Commun Technol Informat Secur & Networking Res Melaka Malaysia Telkom Univ Informat Dept Bandung Indonesia

Currently, the era where social media can present various facilities can answer the needs of the community for information and utilization for socio-economic interests. But the other impact of the presence of social media opens an ample space for the existence of information or hoax news about an event that is troubling the public. The hoax also provides cynical provocation, which is inciting hatred, anger, incitement to many people, directly influencing behavior so that it responds as desired by the hoax makers. Fake news is playing an increasingly dominant role in spreading Misinformation by influencing people's Perceptions or knowledge to distort their awareness and decision-making. A framework is develope dataset collection of hoax gathered using web crawlers from several websites, using classification techniques. This hoax news will be categorized into several detection parameters including, page URL, title hoax news, publish date, author, and content. Matching each word hoax using the similarity algorithm to produce the accuracy of the hoax news uses the rule-based detection method. Experiments were carried out on eleven thousand-hoax news used as training datasets and testing data sets;this data set for validation using similarity algorithms, to produce the highest accuracy of hoax text similarity. In this study, each hoax news will label into four categories, namely, Fact, Hoax, Information, Unknown. Contributions propose Automatic detection of hoax news, Automatic Multilanguage Detection, and a collection of datasets that we gather ourselves and validation that results in four categories of hoax news that have measured in terms of text similarity using similarity techniques. Further research can be continued by adding objects hate speech, black campaign, blockchain technique to ward off hoaxes, or can produce algorithms that produce better text accuracy.

关键词： Component hoax news framework web crawling detection multilanguage unsupervised algorithm similarity algorithm

来源：评论

学校读者我要写书评

暂无评论

Study of Subjective Quality and Objective Blind Quality Prediction of Stereoscopic Videos

引用

IEEE TRANSACTIONS ON IMAGE PROCESSING 2019年第10期28卷 5027-5040页

作者： Appina, Balasubramanyam Dendi, Sathya Veera Reddy Manasa, K. Channappayya, Sumohana S. Bovik, Alan C. IIT Hyderabad Lab Video & Image Anal LFOVIA Dept Elect Engn Kandi 502285 India Univ Texas Austin Dept Elect & Comp Engn LIVE Austin TX 78712 USA

We present a new subjective and objective study on full high-definition (HD) stereoscopic (3D or S3D) video quality. In the subjective study, we constructed an S3D video dataset with 12 pristine and 288 test videos, and the test videos are generated by applying the H.264 and H.265 compression, blur, and frame freeze artifacts. We also propose a no reference (NR) objective video quality assessment (QA) algorithm that relies on measurements of the statistical dependencies between the motion and disparity subband coefficients of S3D videos. Inspired by the Generalized Gaussian Distribution (GGD) approach, we model the joint statistical dependencies between the motion and disparity components as following a Bivariate Generalized Gaussian Distribution (BGGD). We estimate the BGGD model parameters (alpha, beta) and the coherence measure (Psi) from the eigenvalues of the sample covariance matrix (M) of the BCGD. In turn, we model the BGGD parameters of pristine S3D videos using a Multivariate Gaussian (MVG) distribution. The likelihood of a test video's MVG model parameters coming from the pristine MVG model is computed and shown to play a key role in the overall quality estimation. We also estimate the global motion content of each video by averaging the SSIM scores between pairs of successive video frames. To estimate the test S3D video's spatial quality, we apply the popular 2D NR unsupervised NIQE image QA model on a frame-by-frame basis on both views. The overall quality of a test S3D video is finally computed by pooling the test S3D video's likelihood estimates, global motion strength, and spatial quality scores. The proposed algorithm, which is completely blind (requiring no reference videos or training on subjective scores) is called the Motion and Disparity-based 3D video quality evaluator (MoDi(3D)). We show that MoDi(3D) delivers competitive performance over a wide variety of datasets, including the IRCCYN dataset, the WaterlooIVC Phase I dataset, the LFOVIA dat

关键词： Stereoscopic video subjective study full-HD unsupervised algorithm joint statistics

来源：评论

学校读者我要写书评

暂无评论

Modeling the learner within an ILE: Attribute scoring and unsupervised classification of the profile 19

Modeling the learner within an ILE: Attribute scoring and un...

引用

2nd Conference of the Moroccan-Classification-Society (SMC) on New Challenges in Data Sciences - Acts

作者： Ouali, Yassine Oumaira, Ilham Rochdi, Youssef Ibn Tofail Univ EECOMAS Kenitra Morocco

ISBN: (纸本)9781450361293

The observation and the monitoring of the learner is a fundamental problem for interactive learning environments (ILE). The ILE needs to provide tools to replace traditional means and the analysis of the learner's interactions with his learning environment must fill this gap through digital traces. Thus, we propose in our research, an approach of identification, analysis and interpretation of the traces allowing all the actors of the ILE (including the learner) to improve the experience of the learner, throughout his learning cycle. This approach is based on the implementation of attributes associated with the learner to which we will assign scores, as well as iterative clustering algorithms for modeling the learner profile. In addition, it is based on the adaptation of the course according to this profile in order to provide course recommendations. We will use the data sources offered by our MOOC "Back to English" (Ibn Tofail University, Kenitra, Morocco) to illustrate our modeling. In this article, we will focus only on the learner modeling steps using scoring techniques applied to collected traces as well as the grouping of learners via unsupervised clustering algorithms.

关键词： MOOC ILE learners resources course trace analysis scoring unsupervised algorithm clustering

来源：评论

学校读者我要写书评

暂无评论

Machine Learning in the service of learning paths in Interactive Learning Environments

Machine Learning in the service of learning paths in Interac...

引用

International Symposium on Advanced Electrical and Communication Technologies (ISAECT)

作者： Ouali, Yassine Ouamaira, Ilham Rochdi, Youssef Ibn Tofail Univ EECOMAS Kenitra Morocco

ISBN: (纸本)9781538673287

The observation and monitoring of the learner is a fundamental problem facing online learning via Interactive Learning Environments (ILE). The ILE must afford tools to replace traditional means and the analysis of the learner's interactions with his interactive learning environment must fill this gap through the digital traces. Thereby, we propose in this article an approach of identification, analysis and interpretation of the traces allowing all the actors of the ILE (the learner included) to improve the experience of the learner, throughout its learning cycle. This approach is based on the modelization of the learner profile through the implementation of attributes and indicators at all levels of interactions, as well as iterative algorithms of unsupervised classification (clustering). In addition, it is based on the adaptation of the path according to the cluster of each learner based on association algorithms and a recommendation system. We will use the Massive Open MOOC and we will rely on the data sources offered by our MOOC 'Back to English' (Ibn Tofail University, Kenitra, Morocco) deployed on the MOODLE platform to illustrate our modeling.

关键词： MOOC ILE learners ressources path traces analysis indicators unsupervised algorithm clustering recommendations system

来源：评论

学校读者我要写书评

暂无评论

Research on Discovery and Classification Technology of Electric Power Marketing Field Terminals

Research on Discovery and Classification Technology of Elect...

引用

作者： Xianzhou Gao Ruxia Yang Wei Chen Congcong Shi Global Energy Interconnection Research Institute State Grid Key Laboratory of Information & Network Security

In view of the characteristics of diversity,openness,complexity of Electric Power Marketing Field *** may be some security risks such as illegal terminal *** the problem of discovery and classification of Electric Power Marketing Field Terminals is need to be solved,and then we can identify the types of illegal terminals in time and take corresponding *** paper aims at the diversity of Electric Power Marketing Field Terminals and their own differentiation characteristics,and proposes a technology without *** installing client,it can automatically realize terminal discovery,and so solve the traditional problem of non-agent terminal *** the same time,through K-means clustering algorithm terminal model identification,using unsupervised algorithm to extract and identify terminal type fingerprint information,it can achieve accurate classification of terminals,and provide timely alarm information and equipment data for network control and security protection.

关键词： unsupervised algorithm K-means clustering algorithm non-agent accurate classification terminal type fingerprint

来源：评论

学校读者我要写书评

暂无评论

An unsupervised algorithm for the induction of constructions

An unsupervised algorithm for the induction of constructions

引用

IEEE International Conference on Development and Learning (ICDL)

作者： Gaspers, Judith Cimiano, Philipp Griffiths, Sascha S. Wrede, Britta Univ Bielefeld Cognit Interact Technol Excellence Ctr Semant Comp Grp D-4800 Bielefeld Germany Univ Bielefeld Cognit Interact Technol Excellence Ctr Appl Informat D-4800 Bielefeld Germany

ISBN: (纸本)9781612849904

We present an approach to the unsupervised induction of constructions for a specific domain. The main features of our approach are that i) it does not require any supervision in the form of explicit tutoring, ii) it learns pairings between form and meaning and iii) it induces complex syntactic constructions (and their arguments) together with a mapping to a semantic representation beyond mere word-concept associations. A comparison to approaches from the area of learning grammars for semantic parsing shows that the results of our approach are indeed competitive.

关键词： complex syntactic constructions explicit tutoring grammar learning grammars semantic parsing semantic representation unsupervised algorithm unsupervised learning word-concept associations

来源：评论

学校读者我要写书评

暂无评论

A novel methodology for identifying environmental exposures using GPS data

引用

INTERNATIONAL JOURNAL OF GEOGRAPHICAL INFORMATION SCIENCE 2016年第10期30卷 1944-1960页

作者： Cetateanu, Andreea Luca, Bogdan-Alexandru Popescu, Andrei Alin Page, Angie Cooper, Ashley Jones, Andy Univ E Anglia Sch Environm Sci Norwich NR4 7TJ Norfolk England Univ E Anglia Sch Comp Sci Norwich NR4 7TJ Norfolk England Univ Bristol Sch Policy Studies Ctr Exercise Nutr & Hlth Sci Bristol Avon England Univ E Anglia Norwich Med Sch Norwich NR4 7TJ Norfolk England

Aim: While studies using global positioning systems (GPS) have the potential to refine measures of exposure to the neighbourhood environment in health research, one limitation is that they do not typically identify time spent undertaking journeys in motorised vehicles when contact with the environment is reduced. This paper presents and tests a novel methodology to explore the impact of this concern. Methods: Using a case study of exposure assessment to food environments, an unsupervised computational algorithm is employed in order to infer two travel modes: motorised and non-motorised, on the basis of which trips were extracted. Additional criteria are imposed in order to improve robustness of the algorithm. Results: After removing noise in the GPS data and motorised vehicle journeys, 82.43% of the initial GPS points remained. In addition, after comparing a sub-sample of trips classified visually of motorised, non-motorised and mixed mode trips with the algorithm classifications, it was found that there was an agreement of 88%. The measures of exposure to the food environment calculated before and after algorithm classification were strongly correlated. Conclusion: Identifying non-motorised exposures to the food environment makes little difference to exposure estimates in urban children but might be important for adults or rural populations who spend more time in motorised vehicles.

关键词： Global positioning systems food environments travel mode unsupervised algorithm

来源：评论

学校读者我要写书评

暂无评论

An unsupervised Framework Towards Sci-Tech Compound Entity Recognition 9th

An Unsupervised Framework Towards Sci-Tech Compound Entity R...

引用

9th International Conference on Knowledge Science, Engineering, and Management (KSEM)

作者： Yan, Yang Liu, Tingwen Guo, Li Zhao, Jiapeng Shi, Jinqiao Natl Engn Lab Informat Secur Technol Beijing Peoples R China Chinese Acad Sci Inst Informat Engn Beijing Peoples R China Univ Chinese Acad Sci Beijing Peoples R China

ISBN: (纸本)9783319476506;9783319476490

Classifying sci-tech compound named entities, such as the names of patents and projects, plays an important role in enhancing many high-level applications. However, there are very little work on this novel and hard problem. Traditional sequence labeling strategies cannot apply on sci-tech compound entities due to heavy cost of human annotation and low data redundancy. This paper concludes three intrinsic characteristics of sci-tech compound entities, and further proposes a generic and unsupervised framework named SCSegVal to address the problem. Our SCSegVal consists of two components: text splitting and segment validating. We reduce the best split of a text to the problem of maximizing the stickiness sum of segments. The construction of indicative words used in segment validating is reduced to the classical minimum set cover problem. Experimental results on classifying real-world science-technology entities show that SCSegVal achieves a sharp increase comparing with the classical supervised HMM-based approach.

关键词： Named entity recognition unsupervised algorithm Segment stickiness Minimum set cover problem

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：