Abstract
Recent studies have revealed that emerging modern machine learning techniques are advantageous to statistical models for text classification, such as SVM. In this study, we discuss the applications of the support vector machine with mixture of kernel (SVM-MK) to design a text classification system. Differing from the standard SVM, the SVM-MK uses the 1-norm based object function and adopts the convex combinations of single feature basic kernels. Only a linear programming problem needs to be resolved and it greatly reduces the computational costs. More important, it is a transparent model and the optimal feature subset can be obtained automatically. A real Chinese corpus from FudanUniversityis used to demonstrate the good performance of the SVM- MK.
Highlights
With the arrival of "information explosion" era, the infinite growth information resources put forward a great challenge to the information processing
Recent studies have revealed that emerging modern machine learning techniques are advantageous to statistical models for text classification, such as support vector machine (SVM)
We discuss the applications of the support vector machine with mixture of kernel (SVM-MK) to design a text classification system
Summary
With the arrival of "information explosion" era, the infinite growth information resources put forward a great challenge to the information processing. Almost all the important machine learning algorithms is introduced in text classification. Joachims introduced SVM method into the text classification [2]. SVM is sensitive to outliers and noises in the training sample and has limited interpretability due to its kernel theory Another problem is that SVM has a high computational complexity because of the solving of large scale quadratic programming in parameter iterative learning procedure. Motivated by above questions and ideas, we propose a new method named support vector machines with mixture of kernel (SVM-MK) to classify the text. In this method the kernel is a convex combination of many finitely basic kernels.
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call