Generalized relevance learning vector quantization
نویسندگان
چکیده
We propose a new scheme for enlarging generalized learning vector quantization (GLVQ) with weighting factors for the input dimensions. The factors allow an appropriate scaling of the input dimensions according to their relevance. They are adapted automatically during training according to the specific classification task whereby training can be interpreted as stochastic gradient descent on an appropriate error function. This method leads to a more powerful classifier and to an adaptive metric with little extra cost compared to standard GLVQ. Moreover, the size of the weighting factors indicates the relevance of the input dimensions. This proposes a scheme for automatically pruning irrelevant input dimensions. The algorithm is verified on artificial data sets and the iris data from the UCI repository. Afterwards, the method is compared to several well known algorithms which determine the intrinsic data dimension on real world satellite image data.
منابع مشابه
Generalized functional relevance learning vector quantization
Generalized learning vector quantization (GRLVQ) is a prototype based classification algorithm with metric adaptation weighting each data dimensions according to their relevance for the classification task. We present in this paper an extension for functional data, which are usually very high dimensional. This approach supposes the data vectors have to be functional representations. Taking into...
متن کاملMetric Adaptation for Optimal Feature Classification in Learning Vector Quantization Applied to Environment Detection
The paper deals with the concept of relevance learning in learning vector quantization. Recent approaches are considered: the generalized learning vector quantization as well as the soft learning vector quantization. It is shown that relevance learning can be included in both methods obtaining similar structured learning rules for prototype learning as well as relevance factor adaptation. We sh...
متن کاملGeneralized Learning Graph Quantization
This contribution extends generalized LVQ, generalized relevance LVQ, and robust soft LVQ to the graph domain. The proposed approaches are based on the basic learning graph quantization (lgq) algorithm using the orbifold framework. Experiments on three data sets show that the proposed approaches outperform lgq and lgq2.1.
متن کاملSparse Functional Relevance Learning in Generalized Learning Vector Quantization
Relevance learning in learning vector quantization is a central paradigm for classi cation task depending feature weighting and selection. We propose a functional approach to relevance learning for highdimensional functional data. For this purpose we compose the relevance pro le by a superposition of only a few parametrized basis functions taking into account the functional character of the dat...
متن کاملClassification of Boar Sperm Head Images using Learning Vector Quantization
We apply Learning Vector Quantization (LVQ) in automated boar semen quality assessment. The classification of single boar sperm heads into healthy (normal) and non-normal ones is based on grey-scale microscopic images only. Sample data was classified by veterinary experts and is used for training a system with a number of prototypes for each class. We apply as training schemes Kohonen’s LVQ1 an...
متن کاملBiomedical Applications of Prototype Based Classifiers and Relevance Learning
In this contribution, prototype-based systems and relevance learning are presented and discussed in the context of biomedical data analysis. Learning Vector Quantization and Matrix Relevance Learning serve as the main examples. After introducing basic concepts and related approaches, example applications of Generalized Matrix Relevance Learning are reviewed, including the classification of adre...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
- Neural networks : the official journal of the International Neural Network Society
دوره 15 8-9 شماره
صفحات -
تاریخ انتشار 2002