Self-Organizing Maps and Unsupervised Classification

Badran, Fouad; Yacoub, Méziane; Thiria, Sylvie

doi:10.1007/3-540-28847-3_7

Cited by 24 publications

(21 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The referent vectors Rvs are determined from a learning data set L statistically representative of the analyzed data set through an iterative learning process. The referent vectors are first initialized to evenly distributed values of the Ra(k) values in the range of the learning set and then computed by minimizing a nonlinear cost function as in the k-means algorithm [Badran et al, 2005]. Then, for each R i of the learning set presented to the SOM, the Euclidian distance between R i and the referent vectors Rvs are computed and the closest referent vector Rv j is selected.…”

Section: The Classification Methodologymentioning

confidence: 99%

“…The topological map was trained according to the procedure described in Kohonen [2001]. The number of neurons was determined empirically from solutions of similar problems and then adjusted as described by Badran et al [2005]. The learning data set L is composed of the Ra(k) daily values for the year 2003, a year in which the cloud coverage was the lowest with respect to the 13 years of observations.…”

Section: The Classification Methodologymentioning

confidence: 99%

See 1 more Smart Citation

Inferring the seasonal evolution of phytoplankton groups in the Senegalo‐Mauritanian upwelling region from satellite ocean‐color spectral measurements

et al. 2015

Self Cite

View full text Add to dashboard Cite

We have investigated the phytoplankton dynamics of the Senegalo‐Mauritanian upwelling region, which is a very productive region, by processing a 13 year set of SeaWiFS satellite ocean‐color measurements using a PHYSAT‐like method. We clustered the spectra of the ocean‐color normalized reflectance (reflectance normalized by a reflectance dependent on chlorophyll‐a concentration only) into 10 significant spectral classes using a Self‐Organized Map (SOM) associated with a hierarchical ascendant classification (HAC). By analyzing a 13 year climatology of these classes, we have been able to outline a coherent scenario describing the Senegalo‐Mauritanian upwelling region in terms of spatiotemporal variability of phytoplankton groups: during the onset of the upwelling (December–February), we mainly observed nanoeukaryote‐type phytoplankton in the coastal area; in April–May, the period corresponding to the maximum chlorophyll‐a concentration, the nanoeukaryote types were replaced by diatom types. This scenario is in agreement with microscope phytoplankton cell observations done during several past cruises.

show abstract

Section: The Classification Methodologymentioning

confidence: 99%

Section: The Classification Methodologymentioning

confidence: 99%

Inferring the seasonal evolution of phytoplankton groups in the Senegalo‐Mauritanian upwelling region from satellite ocean‐color spectral measurements

et al. 2015

Self Cite

View full text Add to dashboard Cite

show abstract

“…That is, unsupervised clustering is performed independently on each class. There are numerous clustering techniques including the K-means [13], fuzzy C-means [14], hierarchical clustering [15], and self-organizing maps [16]; for a detailed review, the reader is referred to [15]. Any of these clustering techniques can be applied in our approach.…”

Section: B Defining the Sample Weightmentioning

confidence: 99%

Efficient SVM training with reduced weighted samples

Nguyen

Phung

Bouzerdoum

2010

The 2010 International Joint Conference on Neural Networks (IJCNN)

View full text Add to dashboard Cite

This paper presents an efficient training approach for support vector machines that will improve their ability to learn from a large or imbalanced data set. Given an original training set, the proposed approach applies unsupervised learning to extract a smaller set of salient training exemplars, which are represented by weighted cluster centers and the target outputs. In subsequent supervised learning, the objective function is modified by introducing a weight for each new training sample and the corresponding penalty term. In this paper, we investigate two methods of defining the weight based on cluster vectors. The proposed SVM training is implemented and tested on two problems: (i) gender classification of facial images using the FERET data set; (ii) income prediction using the UCI Adult Census data set. Experiment results show that compared to standard SVM training, the proposed approach leads to much faster SVM training, produces a more compact classifier while maintaining generalization ability. Abstract-This paper presents an efficient training approach for support vector machines that will improve their ability to learn from a large or imbalanced data set. Given an original training set, the proposed approach applies unsupervised learning to extract a smaller set of salient training exemplars, which are represented by weighted cluster centers and the target outputs. In subsequent supervised learning, the objective function is modified by introducing a weight for each new training sample and the corresponding penalty term. In this paper, we investigate two methods of defining the weight based on cluster vectors. The proposed SVM training is implemented and tested on two problems: (i) gender classification of facial images using the FERET data set; (ii) income prediction using the UCI Adult Census data set. Experiment results show that compared to standard SVM training, the proposed approach leads to much faster SVM training, produces a more compact classifier while maintaining generalization ability.

show abstract

“…One way of dealing with imbalanced data sets is to simply assign the same number of clusters to each class. There exist many clustering techniques including the Kmeans [19], fuzzy C-means [20], hierarchical clustering [21], and self-organizing maps [22]; for a detailed review, the reader is referred to [21]. Although any of the aforementioned clustering techniques can be used, a suitable clustering is usually application-dependent and could be guided by the probability distribution of the input data.…”

Section: The New Supervised Learning Approachmentioning

confidence: 99%

Efficient supervised learning with reduced training exemplars

Nguyen

Bouzerdoum

Phung

2008

2008 IEEE International Joint Conference on Neural Networks (IEEE World Congress on Computational Intelligence)

View full text Add to dashboard Cite

In this article, we propose a new supervised learning approach for pattern classification applications involving large or imbalanced data sets. In this approach, a clustering technique is employed to reduce the original training set into a smaller set of representative training exemplars, represented by weighted cluster centers and their target outputs. Based on the proposed learning approach, two training algorithms are derived for feedforward neural networks. These algorithms are implemented and tested on two pattern classification applications -skin detection and image classification. Experimental results show that with the proposed learning approach, it is possible to design networks in a fraction of time taken by the standard learning approach, without compromising the generalization ability and overall classification performance. Abstract-In this article, we propose a new supervised learning approach for pattern classification applications involving large or imbalanced data sets. In this approach, a clustering technique is employed to reduce the original training set into a smaller set of representative training exemplars, represented by weighted cluster centers and their target outputs. Based on the proposed learning approach, two training algorithms are derived for feed-forward neural networks. These algorithms are implemented and tested on two pattern classification applications -skin detection and image classification. Experimental results show that with the proposed learning approach, it is possible to design networks in a fraction of time taken by the standard learning approach, without compromising the generalization ability and overall classification performance. Disciplines Computer Sciences | Physical Sciences and Mathematics

show abstract

Self-Organizing Maps and Unsupervised Classification

Cited by 24 publications

References 12 publications

Inferring the seasonal evolution of phytoplankton groups in the Senegalo‐Mauritanian upwelling region from satellite ocean‐color spectral measurements

Inferring the seasonal evolution of phytoplankton groups in the Senegalo‐Mauritanian upwelling region from satellite ocean‐color spectral measurements

Efficient SVM training with reduced weighted samples

Efficient supervised learning with reduced training exemplars

Contact Info

Product

Resources

About