检索结果-内蒙古大学图书馆

IEEE workshop on machine learning for Signal Processing

作者： Frank Nielsen Ke Sun Sony Computer Science Laboratories Inc. Japan Data61 CSIRO Australia

The total variation distance is a core statistical distance between probability measures that satisfies the metric axioms, with value always falling in [0,1]. Since the total variation distance does not admit closed-form expressions for statistical mixtures, one often has to rely in practice on costly numerical integrations or on fast Monte Carlo approximations that however do not guarantee deterministic bounds. In this work, we consider two methods for bounding the total variation of univariate mixture models: The first is based on the information monotonicity property of the total variation to design guaranteed nested deterministic lower bounds. The second method relies on computing the geometric lower and upper envelopes of weighted mixture components to derive deterministic bounds based on density ratio. We demonstrate the tightness of our bounds through simulating Gaussian, Gamma and Rayleigh mixture models.

关键词： TV Upper bound Mixture models Density measurement Probability Monte Carlo methods

来源：评论

学校读者我要写书评

暂无评论

Semantic web: learning from machine learning

Semantic web: Learning from machine learning

引用

2017 Joint international workshops on Hybrid statistical Semantic Understanding and Emerging Semantics, and Semantic Statistics, HybridSemStats 2017

作者： Brickley, Dan Google United States

Recent breakthroughs in the machine learning community have sometimes been characterized as a kind of steamroller. Having first crushed earlier results in image understanding, deep learning researchers have also been making impressive progress with both audio and video understanding and generation, and have had high profile and evocative results with natural language processing. While it is not surprising if this trend can seem threatening (in the examples given, it threatens to overturn previous methods) to a number of traditional subfields of AI (and their associated approaches), Semantic Web researchers and advocates would be ill-advised to consider deep learning a rival or threatening effort. In fact, we argue that there is much to learn from the success of these methods. © 2017 CEUR-WS. All rights reserved.

关键词： Semantic Web

来源：评论

学校读者我要写书评

暂无评论

Toward Intelligent Detection Modelling for Adversarial Samples in Convolutional Neural Networks

Toward Intelligent Detection Modelling for Adversarial Sampl...

引用

IEEE international workshop on Computer-Aided Modeling, Analysis, and Design of Communication Links and Networks, CAMAD

作者： Zhuobiao Qiao Mianxiong Dong Kaoru Ota Jun Wu School of Electronic Information and Electrical Engineering Shanghai Jiao Tong University Shanghai China Department of Information and Electronic Engineering Muroran Institute of Technology Muroran Japan

ISBN: (纸本)9781538661529

Deep Neural Networks (DNNs) are hierarchical nonlinear architectures that have been widely used in artificial intelligence applications. However, these models are vulnerable to adversarial perturbations which add changes slightly and are crafted explicitly to fool the model. Such attacks will cause the neural network to completely change its classification of data. Although various defense strategies have been proposed, existing defense methods have two limitations. first, the discovery success rate is not very high. Second, existing methods depend on the output of a particular layer in a specific learning structure. In this paper, we propose a powerful method for adversarial samples using Large Margin Cosine Estimate(LMCE). By iteratively calculating the large-margin cosine uncertainty estimates between the model predictions, the results can be regarded as a novel measurement of model uncertainty estimation and is available to detect adversarial samples by training using a simple machine learning algorithm. Comparing it with the way in which adversar- ial samples are generated, it is confirmed that this measurement can better distinguish hostile disturbances. We modeled deep neural network attacks and established defense mechanisms against various types of adversarial attacks. Classifier gets better performance than the baseline model. The approach is validated on a series of standard datasets including MNIST and CIFAR -10, outperforming previous ensemble method with strong statistical significance. Experiments indicate that our approach generalizes better across different architectures and attacks.

关键词： Uncertainty Manifolds Neural networks Perturbation methods Computational modeling Training Speech recognition

来源：评论

学校读者我要写书评

暂无评论

The LUCID-Timepix spacecraft payload and the CERN@school educational programme

引用

JOURNAL OF INSTRUMENTATION 2018年第10期13卷 C10004-C10004页

作者： Hatfield, P. Furnell, W. Shenoy, A. Fox, E. Parker, R. Thomas, L. Inst Res Sch Wellcome Wolfson Bldg165 Queens Gate London SW7 5HD England Univ Oxford Clarendon Lab Parks Rd Oxford OX1 3PU England Univ Kent Sch Comp Canterbury CT2 7NF Kent England Langton Star Ctr Langton LaneNackington Rd Canterbury CT4 7AS Kent England

The Langton Ultimate Cosmic ray Intensity Detector (LUCID) is a payload onboard the satellite TechDemoSat-1, used to study the radiation environment in Low Earth Orbit (similar to 635 km). LUCID operated from 2014 to 2017, collecting over 2.1 million frames of radiation data from its five Timepix detectors on board. LUCID is one of the first uses of the Timepix detector technology in open space, with the data providing useful insight into the performance of this technology in new environments. The data has been analysed using computing resources provided by GridPP, with a novel machine learning algorithm. For managing the LUCID data, we have developed an online platform called Timepix Analysis Platformat School (TAPAS). This provides a swift and simple way for users to analyse data that they collect using Timepix detectors from both LUCID and other school based Timepix projects. These projects constitute a secondary school programme 'CERN@school' that give a framework for novel implementations of conventional classroom experiments using Timepix, as well as letting students contribute to large international scientific collaborations and devise their own research projects.

关键词： Space instrumentation Analysis and statistical methods Radiation monitoring

来源：评论

学校读者我要写书评

暂无评论

Hybrid techniques for knowledge-based NLP: Knowledge graphs meet machine learning and all their friends 9

Hybrid techniques for knowledge-based NLP: Knowledge graphs ...

引用

9th workshops and Tutorials of the international Conference on Knowledge Capture, K-CAP2017

作者： Gomez-Perez, Jose Manuel Vila, Daniel Denaux, Ronald Badenes, Carlos Expert System Madrid Spain Recogn AI Madrid Spain Universidad Politecnica de Madrid Madrid Spain

Many different artificial intelligence techniques can be used to explore and exploit large document corpora that are available inside organizations and on the Web. While natural language is symbolic in nature and the first approaches in the field were based on symbolic and rule-based methods, like ontologies, semantic networks and knowledge bases, most widely used methods are currently based on statistical approaches, including linear methods, such as support vectors machines or probabilistic topic models, and nonlinear ones such as neural networks. Each of these two main schools of thought in natural language processing, knowledge-based and statistical, have their limitations and strengths and there is an increasing trend that seeks to combine them in complementary ways to get the best of both worlds. This tutorial will cover the foundations and modern practical applications of knowledge-based and statistical methods, techniques and models and their combination for exploiting large document corpora. The tutorial will first focus on the foundations of many of the techniques that can be used to this purpose, including knowledge graphs, word embeddings, neural network methods, and probabilistic topic models, and will then show how these techniques are being effectively combined in practical applications, including commercial projects where the instructors currently participate. © 2017 Copyright held by the owner/author(s).

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

learning influential genes on cancer gene expression data with stacked denoising autoencoders

Learning influential genes on cancer gene expression data wi...

引用

Biological Ontologies and Knowledge Bases workshop at IEEE international Conference on Bioinformatics and Biomedicine (IEEE BIBM)

作者： Teixeira, Vitor Camacho, Rui Ferreira, Pedro G. Univ Porto Fac Engn MIEIC Porto Portugal Univ Porto Fac Engn DEI Porto Portugal Univ Porto Inst Invest & Inovacao Saude I3S Ipatimup Inst Mol Pathol & Immunol Porto Portugal

ISBN: (纸本)9781509030507

Cancer genome projects are characterizing the genome, epigenome and transcriptome of a large number of samples using the latest high-throughput sequencing assays. The generated data sets pose several challenges for traditional statistical and machine learning methods. In this work we are interested in the task of deriving the most informative genes from a cancer gene expression data set. For that goal we built denoising autoencoders (DAE) and stacked denoising autoencoders and we studied the influence of the input nodes on the final representation of the DAE. We have also compared these deep learning approaches with other existing approaches. Our study is divided into two main tasks. first, we built and compared the performance of several feature extraction methods as well as data sampling methods using classifiers that were able to distinguish the samples of thyroid cancer patients from samples of healthy persons. In the second task, we have investigated the possibility of building comprehensible descriptions of gene expression data by using Denoising Autoencoders and Stacked Denoising Autoencoders as feature extraction methods. After extracting information related to the description built by the network, namely the connection weights, we devised post-processing techniques to extract comprehensible and biologically meaningful descriptions out of the constructed models. We have been able to build high accuracy models to discriminate thyroid cancer from healthy patients but the extraction of comprehensible models is still very limited.

关键词： Deep learning Gene Expression Analysis Knowledge Extraction RNA-Seq Cancer

来源：评论

学校读者我要写书评

暂无评论

Risk prediction using common and rare genetic variants: application to Type 2 diabetes

Risk prediction using common and rare genetic variants: appl...

引用

Biological Ontologies and Knowledge Bases workshop at IEEE international Conference on Bioinformatics and Biomedicine (IEEE BIBM)

作者： Bae, Sunghwan Park, Taesung Seoul Natl Univ Interdisciplinary Program Bioinformat Seoul South Korea Seoul Natl Univ Dept Stat Seoul South Korea

ISBN: (纸本)9781509030507

The success of genome-wide association studies and the development of next generation sequencing technology have identified many disease-related common and rare variants. There have been several prediction models suggested using penalized regression or statistical learning methods. However, only a few prediction models are available which use both common and rare variants. The aim of our study is to compare the performance of prediction models systematically by using both common and rare variants. We first constructed risk prediction models, such as stepwise linear regression, least absolute shrinkage and selection operator, Elastic-Net and support vector machine. We then compared prediction accuracy by calculating the area under the curve. Our results show that the performance using both common and rare variants was better than using either the common variants only or the rare variants only.

关键词： Whole exome sequencing (WES) Risk prediction model Type 2 diabetes (T2D) Penalized regression methods Stepwise selection Support vector machine (SVM)

来源：评论

学校读者我要写书评

暂无评论

Wind speed forecasting using statistical and machine learning methods: A case study in the UAE 1

引用

5th international workshop on Data Analytics for Renewable Energy Integration, DARE 2017

作者： Al Dhaheri, Khawla Woon, Wei Lee Aung, Zeyar Department of Computer Science Khalifa University of Science and Technology Masdar Institute Abu Dhabi United Arab Emirates

ISBN: (数字)9783319716435

ISBN: (纸本)9783319716428

Wind energy is a source of sustainable energy which is developing very quickly all over the world. Forecasting wind speed is a global concern and a critical issue for wind power conversion systems as it has a great influence in the scheduling of power systems as well as on the dynamic control of wind turbines. In this research, we deploy and study four forecasting models in order to forecast wind speeds in the city of Abu Dhabi, United Arab Emirates (UAE). Two of these models are conventional statistical methods, namely, (i) Auto Regression Integrated Moving Average (ARIMA) and (ii) Seasonal Auto Regression Integrated Moving Average (SARIMA) models, and the other two are drawn from the field of machine learning, namely, (i) Artificial Neural Networks (ANN) and (ii) Singular Spectrum Analysis (SSA) models. We compare the performances of these four models in order to determine the model which is most effective for forecasting wind speed data. The results show that the forecasting model SSA provides, on average, the most accurate forecasted values compared to the other three models. However, those three models, ARIMA, SARIMA and ANN, offer better results for the first few hours (around 24 h), which indicates that ARIMA, SARIMA, and ANN models are suitable for short-term forecasting, while SSA is suitable for long-term forecasting. The findings of our research could contribute in defining the fitting forecasting model in terms of short-term forecasting or long-term forecasting. © Springer international Publishing AG 2017.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

An Improved Convergence Analysis of Cyclic Block Coordinate Descent-type methods for Strongly Convex Minimization 19

An Improved Convergence Analysis of Cyclic Block Coordinate ...

引用

19th international Conference on Artificial Intelligence and Statistics (AISTATS)

作者： Li, Xingguo Zhao, Tuo Arora, Raman Liu, Han Hong, Mingyi Univ Minnesota Minneapolis MN 55455 USA Johns Hopkins Univ Baltimore MD 21218 USA Princeton Univ Princeton NJ 08544 USA Iowa State Univ Ames IA USA

The cyclic block coordinate descent-type (CBCD-type) methods have shown remarkable computational performance for solving strongly convex minimization problems. Typical applications include many popular statistical machine learning methods such as elastic-net regression, ridge penalized logistic regression, and sparse additive regression. Existing optimization literature has shown that the CBCD-type methods attain iteration complexity of O(p . log(1/epsilon)), where epsilon is a pre-specified accuracy of the objective value, and p is the number of blocks. However, such iteration complexity explicitly depends on p, and therefore is at least p times worse than those of gradient descent methods. To bridge this theoretical gap, we propose an improved convergence analysis for the CBCD-type methods. In particular, we first show that for a family of quadratic minimization problems, the iteration complexity of the CBCD-type methods matches that of the GD methods in term of dependency on p (up to a log 2 p factor). Thus our complexity bounds are sharper than the existing bounds by at least a factor of p/log(2) p. We also provide a lower bound to confirm that our improved complexity bounds are tight (up to a log(2) p factor) if the largest and smallest eigen-values of the Hessian matrix do not scale with p. Finally, we generalize our analysis to other strongly convex minimization problems beyond quadratic ones.

关键词： Artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

A deterministic Partition Function Approximation for Exponential Random Graph Models 24

A Deterministic Partition Function Approximation for Exponen...

引用

1st international workshop on Social Influence Analysis / 24th international Joint Conference on Artificial Intelligence (IJCAI)

作者： Pu, Wen Choi, Jaesik Hwang, Yunseong Amir, Eyal LinkedIn Corp Mountain View CA 94043 USA Ulsan Natl Inst Sci & Technol Ulsan South Korea Univ Illinois Urbana IL USA

ISBN: (纸本)9781577357384

Exponential Random Graphs Models (ERGM) are common, simple statistical models for social network and other network structures. Unfortunately, inference and learning with them is hard even for small networks because their partition functions are intractable for precise computation. In this paper, we introduce a new quadratic time deterministic approximation to these partition functions. Our main insight enabling this advance is that subgraph statistics is sufficient to derive a lower bound for partition functions given that the model is not dominated by a few graphs. The proposed method differs from existing methods in its ways of exploiting asymptotic properties of subgraph statistics. Compared to the current Monte Carlo simulation based methods, the new method is scalable, stable, and precise enough for inference tasks.

关键词： Monte Carlo methods

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：