In this demo, I use the EM algorithm with a Rauch-Tung-Striebel smoother and an M step, which I ve recently derived, to train a two-layer perceptron, so as to classify medical data (kindly provided by Steve Roberts and Will Penny from EE, Imperial College). The data and simulations are described in: Nando de Freitas, Mahesan Niranjan and Andrew Gee Nonlinear State Space Estimation with Neural Networks and the EM algorithm After downloading the file, type "tar -xf EMdemo.tar" to uncompress it. This creates the directory EMdemo containing the required m files. Go to this directory, load matlab5 and type "EMtremor". The figures will then show you the simulation results, including ROC curves, likelihood plots, decision boundaries with error bars, etc. WARNING: Do make sure that you monitor the log-likelihood and check that it is increasing. Due to numerical errors, it might show glitches for some data sets.
标签： Rauch-Tung-Striebel algorithm smoother which
上传时间： 2016-04-15
上传用户：zhenyushaw
sbgcop: Semiparametric Bayesian Gaussian copula estimation This package estimates parameters of a Gaussian copula, treating the univariate marginal distributions as nuisance parameters as described in Hoff(2007). It also provides a semiparametric imputation procedure for missing multivariate data. Version: 0.95 Date: 2007-03-09 Author: Peter Hoff Maintainer: Peter Hoff <hoff at stat.washington.edu> License: GPL Version 2 or later URL: http://www.stat.washington.edu/hoff CRAN checks: sbgcop results Downloads: Package source: sbgcop_0.95.tar.gz MacOS X binary: sbgcop_0.95.tgz Windows binary: sbgcop_0.95.zip Reference manual: sbgcop.pdf
标签： Semiparametric estimation parameters estimates
上传时间： 2016-04-15
上传用户：talenthn
sbgcop: Semiparametric Bayesian Gaussian copula estimation This package estimates parameters of a Gaussian copula, treating the univariate marginal distributions as nuisance parameters as described in Hoff(2007). It also provides a semiparametric imputation procedure for missing multivariate data. Version: 0.95 Date: 2007-03-09 Author: Peter Hoff Maintainer: Peter Hoff <hoff at stat.washington.edu> License: GPL Version 2 or later URL: http://www.stat.washington.edu/hoff CRAN checks: sbgcop results Downloads: Windows binary: sbgcop_0.95.zip
标签： Semiparametric estimation parameters estimates
上传时间： 2016-04-15
上传用户：qilin
sbgcop: Semiparametric Bayesian Gaussian copula estimation This package estimates parameters of a Gaussian copula, treating the univariate marginal distributions as nuisance parameters as described in Hoff(2007). It also provides a semiparametric imputation procedure for missing multivariate data. Version: 0.95 Date: 2007-03-09 Author: Peter Hoff Maintainer: Peter Hoff <hoff at stat.washington.edu> License: GPL Version 2 or later URL: http://www.stat.washington.edu/hoff CRAN checks: sbgcop results Downloads: Reference manual: sbgcop.pdf
标签： Semiparametric estimation parameters estimates
上传时间： 2014-12-08
上传用户：一诺88
CRM源码This file describes some issues that should be implemented in future and how it should be implemented. Note: it should be implemented exactly how it described below, because another parts of system can rely on it.
标签： should implemented describes issues
上传时间： 2016-04-18
上传用户：lixinxiang
JASML is a java byte code compiler, providing yet another approach to view, write and edit java classes, even without the existence of a java source file - using the java macro instructions, those described in The Java Language Specification.
标签： java providing approach compiler
上传时间： 2016-06-13
上传用户：youth25
Schifra Reed-Solomon Error Correcting Code Library http://www.schifra.com Copyright (c) 2000-2007 Arash Partow, All Rights Reserved. The Schifra Reed-Solomon Error Correcting Code Library and all its components are supplied under the terms of the General Schifra License agreement. The contents of the Schifra Reed-Solomon Error Correcting Code Library and its components may not be copied or disclosed except in accordance with the terms of that agreement. URL: http://www.schifra.com/license.html Parties wanting to use the Schifra Reed-Solomon Error Correcting Code Library and its components within an open source, academic or other noncommercial or not-for-profit environment may do so under the guidelines and in complete accordance with the below attached General Public License (version 2). Under the described terms of "free" use for open source and noncommercial purposes of the Schifra
标签： Reed-Solomon Correcting Copyright Schifra
上传时间： 2016-06-25
上传用户：JasonC
Schifra Reed-Solomon Error Correcting Code Library http://www.schifra.com Copyright (c) 2000-2007 Arash Partow, All Rights Reserved. The Schifra Reed-Solomon Error Correcting Code Library and all its components are supplied under the terms of the General Schifra License agreement. The contents of the Schifra Reed-Solomon Error Correcting Code Library and its components may not be copied or disclosed except in accordance with the terms of that agreement. URL: http://www.schifra.com/license.html Parties wanting to use the Schifra Reed-Solomon Error Correcting Code Library and its components within an open source, academic or other noncommercial or not-for-profit environment may do so under the guidelines and in complete accordance with the below attached General Public License (version 2). Under the described terms of "free" use for open source and noncommercial purposes of the Schifra
标签： Reed-Solomon Correcting Copyright Schifra
上传时间： 2014-01-04
上传用户：ljmwh2000
Coaxial feed structures are widely used in ultra-wide band antennas . This paper modeled the characteristic of the monopole antenna feeded by coaxial line by FDTD in the time-domiain,which showes that . Firstly, it introduced the theory of the arithmetic and the particularly realization in the calculation then it described the use in the time-domain finally it analysed several characteristics of the monopole antenna. The arithmetic used in the microstrip antenna is also a quick and economical way to design the antenna.
标签： structures ultra-wide antennas Coaxial
上传时间： 2016-06-28
上传用户：朗朗乾坤
Recent advances in experimental methods have resulted in the generation of enormous volumes of data across the life sciences. Hence clustering and classification techniques that were once predominantly the domain of ecologists are now being used more widely. This book provides an overview of these important data analysis methods, from long-established statistical methods to more recent machine learning techniques. It aims to provide a framework that will enable the reader to recognise the assumptions and constraints that are implicit in all such techniques. Important generic issues are discussed first and then the major families of algorithms are described. Throughout the focus is on explanation and understanding and readers are directed to other resources that provide additional mathematical rigour when it is required. Examples taken from across the whole of biology, including bioinformatics, are provided throughout the book to illustrate the key concepts and each technique’s potential.
标签： experimental generation advances enormous
上传时间： 2016-10-23
上传用户：wkchong
