Category ranking provides a way to classify plain text documents into a pre-determined set of categories. This work proposes to have a look at typical document collections and ana...
In this work, we suggest a new feature selection technique that lets us use the wrapper approach for finding a well suited feature set for distinguishing experiment classes in hig...
Usage of statistical classifiers, namely AdaBoost and its modifications, in object detection and pattern recognition is a contemporary and popular trend. The computatiponal perfor...
Abstract. In this paper, we propose a probabilistic approach to feature selection for multi-class text categorization. Specifically, we regard document class and occurrence of eac...
Ke Wu, Bao-Liang Lu, Masao Uchiyama, Hitoshi Isaha...
A currently popular trend in object detection and pattern recognition is usage of statistical classifiers, namely AdaBoost and its modifications. The speed performance of these cla...