Recently, Independent Component Analysis (ICA) has been applied to not only problems of blind signal separation, but also feature extraction of patterns. However, the effectiveness of features extracted by ICA (ICA fe...
详细信息
Recently, Independent Component Analysis (ICA) has been applied to not only problems of blind signal separation, but also feature extraction of patterns. However, the effectiveness of features extracted by ICA (ICA features) has not been verified yet. As one of the reasons, it is considered that ICA features are obtained by increasing their independence rather than by increasing their class separability. Hence, we can expect that high-performance pattern features are obtained by introducing supervisor into conventional ICA algorithms such that the class separability of features is enhanced.. In this work, we propose SICA by maximizing Mahalanobis distance between classes. Moreover, we propose a new distance measure in which each ICA feature is weighted by the power of principal components consisting of the ICA feature. In the recognition experiments, we demonstrate that the better recognition accuracy for two data sets in UCI machinelearning Repository is attained when using features extracted by the proposed SICA.
the proceedings contain 46 papers. the special focus in this conference is on Discovery Science. the topics include: the discovery science project in Japan;a computational philosophy of science perspective;combining i...
ISBN:
(纸本)9783540429562
the proceedings contain 46 papers. the special focus in this conference is on Discovery Science. the topics include: the discovery science project in Japan;a computational philosophy of science perspective;combining information visualization withdatamining;a view modeling language for computational knowledge discovery;computational discovery of communicable knowledge;bounding negative information in frequent sets algorithms;spherical horses and shared toothbrushes;lessons learned from a workshop on scientific and technological thinking;clipping and analyzing news using machinelearning techniques;towards discovery of deep and wide first-order structures;eliminating useless parts in semi-structured documents using alternation counts;constructing approximate informative basis of association rules;passage-based document retrieval as a tool for text mining with user's information needs;automated formulation of reactions and pathways in nuclear astrophysics;an integrated framework for extended discovery in particle physics;assisting model-discovery in neuroendocrinology;a general theory of deduction, induction, and learning;knowledge navigation on visualizing complementary documents;extracting keywords from a document as a small world;divide and conquer machinelearning for a genomics analogy problem;towards a method of searching a diverse theory space for scientific discovery;computational revision of quantitative scientific models;an efficient derivation for elementary formal systems based on partial unification;mining semi-structured data by path expressions;simplified training algorithms for hierarchical hidden Markov models;discovering repetitive expressions and affinities from anthologies of classical Japanese poems and a practical algorithm to find the best episode patterns.
In this paper, methods of choosing a vehicle out of an image are explored. Digital images are taken from a monocular camera. Image processing techniques are applied to each single frame picture to create the feature v...
详细信息
ISBN:
(纸本)0819441937
In this paper, methods of choosing a vehicle out of an image are explored. Digital images are taken from a monocular camera. Image processing techniques are applied to each single frame picture to create the feature vector. Finally the resulting features are used to classify whether there is a car in the picture or not using support vector machines. the result are compared to those obtained using a neural network. A discussion on techniques to enhance the feature vector and th results from bothlearningmachines will be included.
We propose a new model for supervised classification for datamining applications. this model is based on products of trees. the information given by each predictor variable is separately extracted by means of a recur...
详细信息
Support vector machines (SVM) are learning algorithms derived from statistical learningtheory. the SVM approach was originally developed for binary classification problems. In this paper SVM architectures for multi-c...
详细信息
Support vector machines (SVM) are learning algorithms derived from statistical learningtheory. the SVM approach was originally developed for binary classification problems. In this paper SVM architectures for multi-class classification problems are discussed, in particular we consider binary trees of SVMs to solve the multi-class problem. Numerical results for different classifiers on a benchmark data set of handwritten digits are presented.
Support vector machines (SVM) are learning algorithms derived from statistical learningtheory. the SVM approach was originally developed for binary classification problems. In this paper SVM architectures for multi-c...
详细信息
An algorithm for data condensation using support vector machines (SVM's) is presented. the algorithm extracts data points lying close to the class boundaries, which form a much reduced but critical set for classif...
详细信息
the proceedings contain 198 papers. the topics discussed include: where is the intelligence in machine vision?;radial-basis-function networks: learning and applications;dimensionality reduction techniques for multivar...
ISBN:
(纸本)0780364007
the proceedings contain 198 papers. the topics discussed include: where is the intelligence in machine vision?;radial-basis-function networks: learning and applications;dimensionality reduction techniques for multivariate data classification, interactive visualization, and analysis-systematic feature selection vs. extraction;integrating community services- a common infrastructure proposal;communication-support for ongoing conversations in users' background knowledge;change in human behaviors based on affiliation needs - toward the design of a social guide agent system;consumer communication on the net-findings from consumer interviews;learning incremental syntactic structures with recursive neural networks;computational capabilities of linear recursive networks;a generalized regression neural network for logo recognition;and a proposal of fuzzy modeling with dimensionality reduction incorporating fuzzy inference method.
the proceedings contain 198 papers. the topics discussed include: where is the intelligence in machine vision?;radial-basis-function networks: learning and applications;dimensionality reduction techniques for multivar...
ISBN:
(纸本)0780364007
the proceedings contain 198 papers. the topics discussed include: where is the intelligence in machine vision?;radial-basis-function networks: learning and applications;dimensionality reduction techniques for multivariate data classification, interactive visualization, and analysis-systematic feature selection vs. extraction;integrating community services- a common infrastructure proposal;communication-support for ongoing conversations in users' background knowledge;change in human behaviors based on affiliation needs - toward the design of a social guide agent system;consumer communication on the net-findings from consumer interviews;learning incremental syntactic structures with recursive neural networks;computational capabilities of linear recursive networks;a generalized regression neural network for logo recognition;and a proposal of fuzzy modeling with dimensionality reduction incorporating fuzzy inference method.
暂无评论