Principal Components Analysis Competitive Learning

2004 ◽  
Vol 16 (11) ◽  
pp. 2459-2481 ◽  
Author(s):  
Ezequiel López-Rubio ◽  
Juan Miguel Ortiz-de-Lazcano-Lobato ◽  
José Muñoz-Pérez ◽  
José Antonio Gómez-Ruiz
Keyword(s):  
Dimensionality Reduction ◽  
Principal Components Analysis ◽  
Principal Components ◽  
Neural Model ◽  
Competitive Learning ◽  
Experimental Results ◽  
Data Set ◽  
Input Distribution ◽  
Local Pca ◽  
Components Analysis

We present a new neural model that extends the classical competitive learning by performing a principal components analysis (PCA) at each neuron. This model represents an improvement with respect to known local PCA methods, because it is not needed to present the entire data set to the network on each computing step. This allows a fast execution while retaining the dimensionality-reduction properties of the PCA. Furthermore, every neuron is able to modify its behavior to adapt to the local dimensionality of the input distribution. Hence, our model has a dimensionality estimation capability. The experimental results we present show the dimensionality-reduction capabilities of the model with multisensor images.

DYNAMIC COMPETITIVE PROBABILISTIC PRINCIPAL COMPONENTS ANALYSIS

2009 ◽  
Vol 19 (02) ◽  
pp. 91-103 ◽  
Author(s):  
EZEQUIEL LÓPEZ-RUBIO ◽  
JUAN MIGUEL ORTIZ-DE-LAZCANO-LOBATO
Keyword(s):  
Principal Components Analysis ◽  
Principal Components ◽  
A Priori ◽  
Image Data ◽  
Neural Model ◽  
Multispectral Image ◽  
Principal Directions ◽  
Local Pca ◽  
Components Analysis ◽  
Basis Vectors

We present a new neural model which extends the classical competitive learning (CL) by performing a Probabilistic Principal Components Analysis (PPCA) at each neuron. The model also has the ability to learn the number of basis vectors required to represent the principal directions of each cluster, so it overcomes a drawback of most local PCA models, where the dimensionality of a cluster must be fixed a priori. Experimental results are presented to show the performance of the network with multispectral image data.

Face Recognition Algorithm Based on Kernel Collaborative Representation

2013 ◽  
Vol 756-759 ◽  
pp. 3590-3595
Author(s):  
Liang Zhang ◽  
Ji Wen Dong
Keyword(s):  
Face Recognition ◽  
Principal Components Analysis ◽  
Principal Components ◽  
Recognition Algorithm ◽  
Experimental Results ◽  
Collaborative Representation ◽  
Nonlinear Structures ◽  
Components Analysis ◽  
Kernel Principal Components Analysis ◽  
The Relationship

Aiming at solving the problems of occlusion and illumination in face recognition, a new method of face recognition based on Kernel Principal Components Analysis (KPCA) and Collaborative Representation Classifier (CRC) is developed. The KPCA can obtain effective discriminative information and reduce the feature dimensions by extracting faces nonlinear structures features, the decisive factor. Considering the collaboration among the samples, the CRC which synthetically consider the relationship among samples is used. Experimental results demonstrate that the algorithm obtains good recognition rates and also improves the efficiency. The KCRC algorithm can effectively solve the problem of illumination and occlusion in face recognition.

scholarly journals Dietary patterns in UK adolescents obtained from a dual-source FFQ and their associations with socio-economic position, nutrient intake and modes of eating

2013 ◽  
Vol 17 (7) ◽  
pp. 1476-1485 ◽  
Author(s):  
Kate Northstone ◽  
Andrew DAC Smith ◽  
Victoria L Cribb ◽  
Pauline M Emmett
Keyword(s):  
Principal Components Analysis ◽  
Principal Components ◽  
Dietary Patterns ◽  
Data Set ◽  
Sociodemographic Differences ◽  
Traditional Health ◽  
Parents And Children ◽  
Components Analysis ◽  
Positive Correlations ◽  
Dietary Data

AbstractObjectiveTo derive dietary patterns using principal components analysis from separate FFQ completed by mothers and their teenagers and to assess associations with nutrient intakes and sociodemographic variables.DesignTwo distinct FFQ were completed by 13-year-olds and their mothers, with some overlap in the foods covered. A combined data set was obtained.SettingAvon Longitudinal Study of Parents and Children (ALSPAC), Bristol, UK.SubjectsTeenagers (n 5334) with adequate dietary data.ResultsFour patterns were obtained using principal components analysis: a ‘Traditional/health-conscious’ pattern, a ‘Processed’ pattern, a ‘Snacks/sugared drinks’ pattern and a ‘Vegetarian’ pattern. The ‘Traditional/health-conscious’ pattern was the most nutrient-rich, having high positive correlations with many nutrients. The ‘Processed’ and ‘Snacks/sugared drinks’ patterns showed little association with important nutrients but were positively associated with energy, fats and sugars. There were clear gender and sociodemographic differences across the patterns. Lower scores were seen on the ‘Traditional/health conscious’ and ‘Vegetarian’ patterns in males and in those with younger and less educated mothers. Higher scores were seen on the ‘Traditional/health-conscious’ and ‘Vegetarian’ patterns in girls and in those whose mothers had higher levels of education.ConclusionsIt is important to establish healthy eating patterns by the teenage years. However, this is a time when it is difficult to accurately establish dietary intake from a single source, since teenagers consume increasing amounts of foods outside the home. Further dietary pattern studies should focus on teenagers and the source of dietary data collection merits consideration.

The use of principal components analysis for the investigation of an organic air pollutants data set

1984 ◽  
Vol 18 (11) ◽  
pp. 2471-2478 ◽  
Author(s):  
J. Smeyers-Verbeke ◽  
J.C. Den Hartog ◽  
W.H. Dehker ◽  
D. Coomans ◽  
L. Buydens ◽  
...  
Keyword(s):  
Principal Components Analysis ◽  
Principal Components ◽  
Air Pollutants ◽  
Data Set ◽  
Components Analysis

scholarly journals Multivariate Characterization of Hydrogen Balmer Emission in Cataclysmic Variables

2006 ◽  
Vol 23 (3) ◽  
pp. 106-118 ◽  
Author(s):  
Gordon E. Sarty ◽  
Kinwah Wu
Keyword(s):  
Principal Components Analysis ◽  
Discriminant Function ◽  
Principal Components ◽  
Discriminant Function Analysis ◽  
Function Analysis ◽  
Cataclysmic Variables ◽  
Data Set ◽  
Analysis Methods ◽  
Wide Range ◽  
Components Analysis

AbstractThe ratios of hydrogen Balmer emission line intensities in cataclysmic variables are signatures of the physical processes that produce them. To quantify those signatures relative to classifications of cataclysmic variable types, we applied the multivariate statistical analysis methods of principal components analysis and discriminant function analysis to the spectroscopic emission data set of Williams (1983). The two analysis methods reveal two different sources of variation in the ratios of the emission lines. The source of variation seen in the principal components analysis was shown to be correlated with the binary orbital period. The source of variation seen in the discriminant function analysis was shown to be correlated with the equivalent width of the Hβ line. Comparison of the data scatterplot with scatterplots of theoretical models shows that Balmer line emission from T CrB systems is consistent with the photoionization of a surrounding nebula. Otherwise, models that we considered do not reproduce the wide range of Balmer decrements, including ‘inverted’ decrements, seen in the data.

scholarly journals A method of dimensionality reduction by selection of components in principal component analysis for text classification

Filomat ◽  
2018 ◽  
Vol 32 (5) ◽  
pp. 1499-1506 ◽  
Author(s):  
Yangwu Zhang ◽  
Guohe Li ◽  
Heng Zong
Keyword(s):  
Principal Component Analysis ◽  
Feature Extraction ◽  
Dimensionality Reduction ◽  
Principal Components Analysis ◽  
Text Classification ◽  
Principal Components ◽  
Principal Component ◽  
Method Of Dimensionality Reduction ◽  
Components Analysis ◽  
Selection Of

Dimensionality reduction, including feature extraction and selection, is one of the key points for text classification. In this paper, we propose a mixed method of dimensionality reduction constructed by principal components analysis and the selection of components. Principal components analysis is a method of feature extraction. Not all of the components in principal component analysis contribute to classification, because PCA objective is not a form of discriminant analysis (see, e.g. Jolliffe, 2002). In this context, we present a function of components selection, which returns the useful components for classification by the indicators of the performances on the different subsets of the components. Compared to traditional methods of feature selection, SVM classifiers trained on selected components show improved classification performance and a reduction in computational overhead.

A Principal Components Analysis Self-Organizing Neural Network Model and Computational Experiment

2013 ◽  
Vol 756-759 ◽  
pp. 3330-3335
Author(s):  
Ji Fu Nong
Keyword(s):  
Neural Network ◽  
Principal Components Analysis ◽  
Network Model ◽  
Neural Network Model ◽  
Principal Components ◽  
Computational Experiment ◽  
Neural Model ◽  
Self Organizing Map ◽  
Components Analysis ◽  
Self Organizing

We propose a new self-organizing neural model that performs principal components analysis. It is also related to the adaptive subspace self-organizing map (ASSOM) network, but its training equations are simpler. Experimental results are reported, which show that the new model has better performance than the ASSOM network.

Planktonic Biomass Trajectories In Lake Ecosystems

1983 ◽  
Vol 40 (10) ◽  
pp. 1752-1760 ◽  
Author(s):  
Michael A. Gates ◽  
Ann P. Zimmerman ◽  
W. Gary Sprules ◽  
Roy Knoechel
Keyword(s):  
Hypothesis Testing ◽  
Principal Components Analysis ◽  
Biomass Allocation ◽  
Principal Components ◽  
Primary Component ◽  
Size Category ◽  
Data Set ◽  
Lake Ecosystems ◽  
The Mean ◽  
Components Analysis

We introduce a method, based on principal components analysis, for studying temporal changes in biomass allocation among 16 size–category compartments of lake plankton. Applied to data from a series of 12 Ontario lakes over three sampling seasons, the technique provides a simple means of visualizing shifts in patterns of biomass allocation, and it allows comparative analyses of biomass fluctuations in different lakes. Each of the primary component axes is interpretable. Furthermore, a large proportion of the variance in both the mean position of a lake and its movement along these axes is interpreted as a function of lake physicochemistry. The analysis also provides weighted scores for use in hypothesis testing which are an improvement over mean biomass values alone, because they take into account the structure of variation in the data set.

scholarly journals A Hybrid Fuzzy-SVM classifier for automated lung diseases diagnosis

2016 ◽  
Vol 22 (4) ◽  
pp. 97-103 ◽  
Author(s):  
Donia Ben Hassen ◽  
Sihem Ben Zakour ◽  
Hassen Taleb
Keyword(s):  
Principal Components Analysis ◽  
Principal Components ◽  
Lung Diseases ◽  
Roc Curves ◽  
Chest Radiographs ◽  
Experimental Results ◽  
Svm Classifier ◽  
Forward Selection ◽  
Svm Classification ◽  
Components Analysis

Abstract A novel scheme for lesions classification in chest radiographs is presented in this paper. Features are extracted from detected lesions from lung regions which are segmented automatically. Then, we needed to eliminate redundant variables from the subset extracted because they affect the performance of the classification. We used Stepwise Forward Selection and Principal Components Analysis. Then, we obtained two subsets of features. We finally experimented the Stepwise/FCM/SVM classification and the PCA/FCM/SVM one. The ROC curves show that the hybrid PCA/FCM/SVM has relatively better accuracy and remarkable higher efficiency. Experimental results suggest that this approach may be helpful to radiologists for reading chest images.

Sign in / Sign up

Export Citation Format

Share Document