Median Variants of LVQ for Optimization of Statistical Quality Measures for Classification of Dissimilarity Data
نویسندگان
چکیده
We consider in this article median variants of the learning vector quantization classifier for classification of dissimilarity data. particularly we are interested in optimization of advanced classification quality measures like sensitivity, specificity or the Fβmeasure. These measures are frequently more appropriate than simple accuracy, in particular, if the training data are imbalanced for the investigated data classes. We present the mathematical theory for this approach based on a generalized ExpectationMaximization-scheme. Machine Learning Reports http://www.techfak.uni-bielefeld.de/∼fschleif/mlr/mlr.html Median Variants of LVQ for Optimization of Statistical Quality Measures for Classi cation of Dissimilarity Data David Nebel and Thomas Villmann University of Applied Sciences Mittweida, Fac. of Mathematics/Natural and Computer Sciences, Computational Intelligence Group, Germany
منابع مشابه
Median-LVQ for Classification of Dissimilarity Data based on ROC-Optimization
In this article we consider a median variant of the learning vector quantization (LVQ) classifier for classification of dissimilarity data. However, beside the median aspect, we propose to optimize the receiver-operating characteristics (ROC) instead of the classification accuracy. In particular, we present a probabilistic LVQ model with an adaptation scheme based on a generalized ExpectationMa...
متن کاملDistance Measures for Prototype Based Classification
The basic concepts of distance based classification are introduced in terms of clear-cut example systems. The classical k-NearestNeigbhor (kNN) classifier serves as the starting point of the discussion. Learning Vector Quantization (LVQ) is introduced, which represents the reference data by a few prototypes. This requires a data driven training process; examples of heuristic and cost function b...
متن کاملComposite Kernel Optimization in Semi-Supervised Metric
Machine-learning solutions to classification, clustering and matching problems critically depend on the adopted metric, which in the past was selected heuristically. In the last decade, it has been demonstrated that an appropriate metric can be learnt from data, resulting in superior performance as compared with traditional metrics. This has recently stimulated a considerable interest in the to...
متن کاملA comprehensive experimental comparison of the aggregation techniques for face recognition
In face recognition, one of the most important problems to tackle is a large amount of data and the redundancy of information contained in facial images. There are numerous approaches attempting to reduce this redundancy. One of them is information aggregation based on the results of classifiers built on selected facial areas being the most salient regions from the point of view of classificati...
متن کاملKernel Regularization and Dimension Reduction
It is often possible to use expert knowledge or other sources of information to obtain dissimilarity measures for pairs of objects, which serve as pseudo-distances between the objects. When dissimilarity information is available as the data, there are two different types of problems of interest. The first is to estimate full position configuration for all objects in a low dimensional space whil...
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:
دوره شماره
صفحات -
تاریخ انتشار 2014