We propose in this contribution a method for l1-regularization in prototype based relevance learning vector quantization (LVQ) for sparse relevance pro les. Sparse relevance pro les in hyperspectral data analysis fade down those spectral bands which are not necessary for classi cation. In particular, we consider the sparsity in the relevance pro le enforced by LASSO optimization. The latter one...

This paper presents an application of Learning Vector Quantization (LVQ) neural network (NN) to Automatic Fingerprint Verification (AFV). The new approach is based on both local (minutiae) and global image features (shape signatures). The matched minutiae are used as reference axis for generating shape signatures which are then digitized to form a feature vector describing the fingerprint. A LV...

Winner-Takes-All (WTA) algorithms offer intuitive and powerful learning schemes such as Learning Vector Quantization (LVQ) and variations thereof, most of which are heuristically motivated. In this article we investigate in an exact mathematical way the dynamics of different vector quantization (VQ) schemes including standard LVQ in simple, though relevant settings. We consider the training fro...

Recent extensions of learning vector quantization (LVQ) to general (dis-)similarity data have paved the way towards LVQ classifiers for possibly discrete, structured objects such as sequences addressed by classical alignment. In this contribution, we propose a metric learning scheme based on this framework which allows for autonomous learning of the underlying scoring matrix according to a give...

In this note we formulate image segmentation as a clustering problem. Feature vectors, extracted from a raw image are clustered into subregions, thereby segmenting the image. A fuzzy generalization of Kohonen learning vector quantization (LVQ) which integrates the Fuzzy cMeans (FCM) model with the learning rate and updating strategies of the LVQ Is used for this task. This network, which segmen...

Initial values of reference vectors have significant influence on recognition accuracy in LVQ. There are several existing techniques, such as SOM and k-means, for setting initial values of reference vectors, each of which has provided some positive results. However, those results are not sufficient for the improvement of recognition accuracy. This study proposes an ACO-used method for initializ...

We present a hybrid face recognition algorithm which is based on the linear discriminant analysis (LDA) improved by a fusion technique and learning vector quantization (LVQ) in the paper. Firstly, the improved LDA is utilized to reduce the sample vector dimension, and then the LVQ classifier is used to recognize human faces. We perform intensive set of simulation experiments and results show th...

Identifying exceptional students for scholarships is an essential part of the admissions process in undergraduate and postgraduate institutions, and identifying weak students who are likely to fail is also important for allocating limited tutoring resources. In this article, we have tried to design an intelligent system which can separate and classify student according to learning factor and pe...

The support vector machine (SVM) constitutes one of the most successful current learning algorithms with excellent classification accuracy in large real-life problems and strong theoretical background. However, a SVM solution is given by a not intuitive classification in terms of extreme values of the training set and the size of a SVM classifier scales with the number of training data. General...

SOM and LVQ algorithms for symbol strings have been introduced and applied to isolatedword recognition, for the construction of an optimal pronunciation dictionary for a given speech recognizer.