One of the principal bottlenecks in applying learning techniques to classification problems is the large amount of labeled training data required. Especially for images and video, ...
Ajay J. Joshi, Fatih Porikli, Nikolaos Papanikolop...
Multiclass classification is an important problem in pattern recognition. Hierarchical SVM classifiers such as DAG-SVM and BHC-SVM are popular in solving multiclass problems. Howe...
Databases are a key technology for molecular biology which is a very data intensive discipline. Since molecular biological databases are rather heterogeneous, unification and data...
In many multiclass learning scenarios, the number of classes is relatively large (thousands,...), or the space and time efficiency of the learning system can be crucial. We invest...
In this paper, we examine the performance of the two policies for keyword selection over standard document corpora of varying properties. While in corpus-based policy a single set ...