检索结果-内蒙古大学图书馆

IEEE International Conference on Communications (IEEE ICC)

作者： Kim, Seonho So, Minji Lee, Namyoon Hong, Songnam Ajou Univ Suwon South Korea POSTECH Pohang South Korea

ISBN: (纸本)9781728123738

We study an uplink multiuser multiple-input multiple-output (MU-MIMO) system with one-bit analog-to-digital converters (ADCs). In this system, we recently proposed a supervised-learning (SL) detector by modeling a non-linear end-to-end system function into a parameterized Bernoulli-like model. Despite its attractive performance, the SL detector requires a large amount of labeled data (i.e., pilot signals) to accurately estimate the parameters of the underlying model, since the amount of the parameters grow exponentially with the number of users. In this paper, we address this problem by presenting a semi-supervised learning (SSL) detector where both pilot signals (i.e., labeled data) and some part of data signals (i.e., unlabeled data) are exploited to estimate the parameters via expectation-maximization (em) algorithm. Simulation results demonstrate that the proposed SSL detector can achieve the performance of the existing SL detector with significantly lower pilot-overhead.

关键词： Massive MIMO one-bit ADC MIMO detection machine learning semi-supervised Learning em algorithm

来源：评论

学校读者我要写书评

暂无评论

The improvement of change detection from multi-date satellite images using the source separation 19

The improvement of change detection from multi-date satellit...

引用

IEEE 19th Mediterranean Microwave Symposium (MMS)

作者： Echtioui, Amira Ben Sassi, Olfa Sellami, Lamia Ben Hamida, Ahmed Natl Sch Engn Sfax Adv Technol Med & Signals Sfax Tunisia Higher Inst Comp Sci Adv Technol Med & Signals Tunis El Manar Tunisia Higher Inst Biotechnol Sfax Adv Technol Med & Signals Sfax Tunisia

ISBN: (纸本)9781728140643

The development of satellites with the strong temporal repetitiveness and development of remote sensing techniques resulted in the advancement of change detection techniques from geospatial imagery. The natural events cause many modifications in the control process of the ecosystems. There is a necessity of using a method capable to map, categorize and monitor areas affected by natural events along time. In this article, a novel methodology of change detection is proposed in order to improve the change detection from multi-date satellite image using the source separation. The results obtained by our methodology are efficient and effective.

关键词： change detection source separation JADE algorithm SOBI algorithm em algorithm classification Post-classification multi-date image

来源：评论

学校读者我要写书评

暂无评论

Component Elimination Strategies to Fit Mixtures of Multiple Scale Distributions 3rd

Component Elimination Strategies to Fit Mixtures of Multiple...

引用

Research School in Statistics and Data Science (RSSDS)

作者： Forbes, Florence Arnaud, Alexis Lemasson, Benjamin Barbier, emmanuel Univ Grenoble Alpes INRIA CNRS Grenoble INPInst Engn F-38000 Grenoble France Univ Grenoble Alpes Inserm U1216 Grenoble Inst Neurosci Grenoble France

ISBN: (纸本)9789811519604;9789811519598

We address the issue of selecting automatically the number of components in mixture models with non-Gaussian components. As a more efficient alternative to the traditional comparison of several model scores in a range, we consider procedures based on a single run of the inference scheme. Starting from an overfitting mixture in a Bayesian setting, we investigate two strategies to eliminate superfluous components. We implement these strategies for mixtures of multiple scale distributions which exhibit a variety of shapes not necessarily elliptical while remaining analytical and tractable in multiple dimensions. A Bayesian formulation and a tractable inference procedure based on variational approximation are proposed. Preliminary results on simulated and real data show promising performance in terms of model selection and computational time.

关键词： Gaussian scale mixture Bayesian analysis Bayesian model selection em algorithm Variational approximation

来源：评论

学校读者我要写书评

暂无评论

Flow-Based Network Tomography Agent for Software Defined Data Center 3

Flow-Based Network Tomography Agent for Software Defined Dat...

引用

3rd International Conference on Recent Advances in Signal Processing, Telecommunications and Computing (IEEE SigTelCom)

作者： Biyar, Elham Dehghan Karanlik, Bahtiyar Canberk, Berk Istanbul Tech Univ Dept Comp Engn Istanbul Turkey Sekom Yazilim AS Istanbul Turkey

ISBN: (纸本)9781538679630

In software defined networking (SDN) data centers, collecting real-time routing information of different traffic flows for tomography of data centers requires an up-to-date knowledge of every link. However, traditional techniques are mostly not covering traffic matrix (TM) estimation for specific traffic types. This paper proposes an approach to construct a tomography agent to estimate and manage traffic flows through the data center network. For this aim, we regenerated the traffic matrix according to three different traffic types (bandwidth-sensitive, delay-sensitive, best-effort) and taking into account delay information obtained from the Markov chain M/G/1 priority queuing method both for link count and traffic matrix. Consequently, after regeneration step, we use Expectation Maximization approach for iteratively estimate the traffic matrix. In addition, regarding to estimated traffic matrix, by using Max-Min fairness method and definition of flow utility function, our simulation results in reduction in end-to-end routing delay and also flow utility enhancement.

关键词： SDN data center Tomography Traffic Matrix Markov Chain Max-Min Fairness em algorithm

来源：评论

学校读者我要写书评

暂无评论

Particle em for Variable Selection

引用

JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION 2018年第524期113卷 1684-1697页

作者： Rockova, Veronika Univ Chicago Booth Sch Business Chicago IL 60637 USA

Despite its long history of success, the em algorithm has been vulnerable to local entrapment when the posterior/likelihood is multi-modal. This is particularly pronounced in spike-and-slab posterior distributions for Bayesian variable selection. The main thrust of this article is to introduce the particle em algorithm, a new population-based optimization strategy that harvests multiple modes in search spaces that present many local maxima. Motivated by nonparametric variational Bayes strategies, particle em achieves this goal by deploying an ensemble of interactive repulsive particles. These particles are geared toward uncharted areas of the posterior, providing a more comprehensive summary of its topography than simple parallel em deployments. A sequential Monte Carlo variant of particle em is also proposed that explores a sequence of annealed posteriors by sampling from a set of mutually avoiding particles. Particle em outputs a deterministic reconstruction of the posterior distribution for approximate fully Bayes inference by capturing its essential modes and mode weights. This reconstruction reflects model selection uncertainty and is supported by asymptotic considerations, which indicate that the requisite number of particles need not be large in the presence of sparsity (when p > n). Supplementary materials for this article are available online.

关键词： em algorithm Ensemble optimization Multimodality Spike-and-slab

来源：评论

学校读者我要写书评

暂无评论

A robust multivariate Birnbaum-Saunders distribution: em estimation

引用

STATISTICS 2018年第2期52卷 321-344页

作者： Romeiro, Renata G. Vilca, Filidor Balakrishnan, N. Univ Estadual Campinas Dept Estat Caixa Postal 6065 Sao Paulo Brazil McMaster Univ Dept Math & Stat Hamilton ON Canada

We propose here a robust multivariate extension of the bivariate Birnbaum-Saunders (BS) distribution derived by Kundu et al. [Bivariate Birnbaum-Saunders distribution and associated inference. J Multivariate Anal. 2010;101:113-125], based on scale mixtures of normal (SMN) distributions that are used for modelling symmetric data. This resulting multivariate BS-type distribution is an absolutely continuous distribution whose marginal and conditional distributions are of BS-type distribution of Balakrishnan et al. [Estimation in the Birnbaum-Saunders distribution based on scalemixture of normals and the em algorithm. Stat Oper Res Trans. 2009;33:171-192]. Due to the complexity of the likelihood function, parameter estimation by direct maximization is very difficult to achieve. For this reason, we exploit the nice hierarchical representation of the proposed distribution to propose a fast and accurate em algorithm for computing the maximum likelihood (ML) estimates of the model parameters. We then evaluate the finite-sample performance of the developed em algorithm and the asymptotic properties of the ML estimates through empirical experiments. Finally, we illustrate the obtained results with a real data and display the robustness feature of the estimation procedure developed here.

关键词： Multivariate Birnbaum-Saunders distribution em algorithm kurtosis maximum likelihood method robust estimation scale mixtures of normal distributions

来源：评论

学校读者我要写书评

暂无评论

Using Noisy Word-Level Labels to Train a Phoneme Recognizer based on Neural Networks by Expectation Maximization 19

Using Noisy Word-Level Labels to Train a Phoneme Recognizer ...

引用

Proceedings of the 2019 8th International Conference on Computing and Pattern Recognition

作者： Chen Li Bo Zhang Shan Huang Zhenhuan Liu College of Computer Science Nankai University Tianjin P.R.China College of Software Nankai University Tianjin P.R.China

ISBN: (纸本)9781450376570

The Connectionist Temporal Classification (CTC) technique can be used to train a neural-network based speech recognizer. When the technique is used to train a phoneme recognizer, it is required that training data should be annotated with phoneme-level labels. This is not feasible if large speech databases are used. One approach to make use of such speech data is to convert the word-level transcriptions into phoneme-level labels, followed by a CTC training. The problem of this approach is that the converted phonemelevel labels may mismatch the audio content of the speech data. This paper uses a probabilistic model to describe the probability of observing the noisy phoneme-level labels given an utterance. The model consists of a neural network which predicts the probability of any phoneme sequence, and another so-called mismatch model to describe the probability of disturbing a phoneme sequence to another. Based on the Expectation-Maximization (em) framework, we propose a training algorithm which can simultaneously learn parameters of the neural-network and the mismatch model. Effectiveness of our method is verified by comparing recognition performance of our method with a conventional training method on TIMIT corpus.

关键词： em algorithm noisy labels CTC Phoneme recognition

来源：评论

学校读者我要写书评

暂无评论

A Flexible Zero-Inflated Poisson Regression Model

A Flexible Zero-Inflated Poisson Regression Model

引用

作者： Eric S. Roemmele University of Kentucky

学位级别：博士

A practical problem often encountered with observed count data is the presence of ex- cess zeros. Zero-inflation in count data can easily be handled by zero-inflated models, which is a two-component mixture of a point mass at zero and a discrete distribu- tion for the count data. In the presence of predictors, zero-inflated Poisson (ZIP) regression models are, perhaps, the most commonly used. However, the fully para- metric ZIP regression model could sometimes be restrictive, especially with respect to the mixing proportions. Taking inspiration from some of the recent literature on semiparametric mixtures of regressions models for flexible mixture modeling, we pro- pose a semiparametric ZIP regression model. We present an "em-like" algorithm for estimation and a summary of asymptotic properties of the estimators. The pro- posed semiparametric models are then applied to a data set involving clandestine methamphetamine laboratories and Alzheimer's disease.

关键词： Bootstrap Count data em algorithm zero-inflation semiparametric model

来源：评论

学校读者我要写书评

暂无评论

Multi-source Landmark Fusion based on Machine Learning 2019

Multi-source Landmark Fusion based on Machine Learning

引用

Proceedings of the 2019 the 9th International Conference on Communication and Network Security

作者： Wen Yang Meijuan Yin Xiaonan Liu Can Wang Shunran Duan PLA Strategic Support Force Information Engineering University Zhengzhou Henan China

ISBN: (纸本)9781450376624

Network entity landmark is the key foundation of IP geolocaiton which plays an important role in network security. Integrating multi-source landmarks to generate a landmark database with high IP coverage and high location accuracy is an important solution for improving the IP geolocation effect. However, the city-level landmarks provide the city name while street-level landmarks provide the latitude and longitude where they located. Owning to their inconsistent format, the state-of-art fusion algorithm cannot effectively integrate the two types of data. Hence, this paper proposes Lusion, a multi-source landmark fusion algorithm. We first extend the IP addresses in the landmark data sources, then model the location data of the two types of landmarks using the landmark location mixture model, and finally use the expectation -maximization algorithm to estimate the location of the landmarks. The simulation experiments on 25 landmark data sources show that the algorithm can effectively integrate the city-level and street-level landmarks from different data sources, and have a significantly better performance than the original data sources in the location accuracy. Furthermore, we evaluate Lusion on real-world datasets, which consists of 7 city-level and 3 street-level landmark data sources, by locating 100 IP addresses in Hong Kong and Zhengzhou respectively. The geolocation results show that Lusion increased the city-level accuracy by at least 8 percentage points compared with the original data sources, and reduces the geolocation error from 3.42 km to 2.92 km based on the best original landmark data set.

关键词： Data fusion Network entity landmarks em algorithm Mixture model IP geolocation

来源：评论

学校读者我要写书评

暂无评论

Expectation Maximization Based FitzHugh-Nagumo Model Identification Under Unknown Gaussian Measurement Noise

Expectation Maximization Based FitzHugh-Nagumo Model Identif...

引用

第32届中国控制与决策会议

作者： Li-Hui Geng Terefe Bayisa Ayele Jin-Cang Liu Brett Ninness Tianjin Key Laboratory of Information Sensing and Intelligent Control School of Automation and Electrical EngineeringTianjin University of Technology and Education School of Electrical Engineering and Computer Science The University of Newcastle

This paper applies a new expectation maximization（em） based identification method to estimate a generic Fitz Hugh-Nagumo（FHN） model under unknown Gaussian measurement *** is well noted that such FHN model is an elementary neuronal dynamics description and plays a significant role in deep understanding and basic analysis for complicated mechanisms of some neural *** from the most existing relevant identification methods,the applied new method is additionally capable of supplying users with variance estimation for the unknown measurement noise corrupting on the membrane potential apart from model parameter *** unknown parameters can be iteratively estimated by a particle smoothing based em algorithm,which consists of an expectation（E） step and a maximization（M） *** joint-state particles are produced by a new particle smoothing algorithm to evaluate an expectation of a log-likelihood function in the E step,and the model parameters and noise variance can then be efficiently optimized by the gradient based methods in the M *** resulting estimations own global convergence for a relatively wide range of parameter ***,good convergence behaviors of the estimated model parameters and noise variance are demonstrated by a numerical simulation for a classic FHN model by using 10 simulation realizations with random initializations varying within ±100% of their respective true values.

关键词： FHN model identification em algorithm Particle smoothing Unknown measurement noise

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：