For many supervised learning problems, we possess prior knowledge about which features yield similar information about the target variable. In predicting the topic of a document, ...
Ted Sandler, John Blitzer, Partha Pratim Talukdar,...
Numerous applications of data mining to scientific data involve the induction of a classification model. In many cases, the collection of data is not performed with this task in m...
The symmetrical decomposition is a powerful method to extract features for image recognition. It reveals the significant discriminative information from the mirror image of symmetr...
This paper describes our experiments in Geographical Information Retrieval (GIR) in the context of our participation in the GeoCLEF 2006 Monolingual English task. The TALPGeoIR sy...
Linear Discriminant Analysis (LDA) is a popular feature extraction technique in statistical pattern recognition. However, it often suffers from the small sample size problem when ...