A DUAL COORDINATE DESCENT ALGORITHM FOR SVMs COMBINED WITH RATIONAL KERNELS

2011 ◽  
Vol 22 (08) ◽  
pp. 1761-1779 ◽  
Author(s):  
CYRIL ALLAUZEN ◽  
CORINNA CORTES ◽  
MEHRYAR MOHRI
Keyword(s):  
Speech Processing ◽  
Large Scale ◽  
Support Vector ◽  
Dramatic Reduction ◽  
Training Time ◽  
Weighted Automata ◽  
Coordinate Descent Algorithm ◽  
Vector Machines ◽  
Dual Coordinate Descent ◽  
Rational Kernels

This paper presents a novel application of automata algorithms to machine learning. It introduces the first optimization solution for support vector machines used with sequence kernels that is purely based on weighted automata and transducer algorithms, without requiring any specific solver. The algorithms presented apply to a family of kernels covering all those commonly used in text and speech processing or computational biology. We show that these algorithms have significantly better computational complexity than previous ones and report the results of large-scale experiments demonstrating a dramatic reduction of the training time, typically by several orders of magnitude.

scholarly journals Selection of Support Vector Candidates Using Relative Support Distance for Sustainability in Large-Scale Support Vector Machines

2020 ◽  
Vol 10 (19) ◽  
pp. 6979
Author(s):  
Minho Ryu ◽  
Kichun Lee
Keyword(s):  
Support Vector Machines ◽  
Quadratic Programming ◽  
Decision Trees ◽  
Programming Problem ◽  
Large Scale ◽  
Classification Performance ◽  
Quadratic Programming Problem ◽  
Support Vector ◽  
Training Time ◽  
Vector Machines

Support vector machines (SVMs) are a well-known classifier due to their superior classification performance. They are defined by a hyperplane, which separates two classes with the largest margin. In the computation of the hyperplane, however, it is necessary to solve a quadratic programming problem. The storage cost of a quadratic programming problem grows with the square of the number of training sample points, and the time complexity is proportional to the cube of the number in general. Thus, it is worth studying how to reduce the training time of SVMs without compromising the performance to prepare for sustainability in large-scale SVM problems. In this paper, we proposed a novel data reduction method for reducing the training time by combining decision trees and relative support distance. We applied a new concept, relative support distance, to select good support vector candidates in each partition generated by the decision trees. The selected support vector candidates improved the training speed for large-scale SVM problems. In experiments, we demonstrated that our approach significantly reduced the training time while maintaining good classification performance in comparison with existing approaches.

Research on Large Scale Data Set Processing Based on SVM

2011 ◽  
Vol 216 ◽  
pp. 738-741
Author(s):  
Yue E Chen ◽  
Bai Li Ren
Keyword(s):  
Large Scale ◽  
Support Vector ◽  
Simulation Experiments ◽  
Data Set ◽  
Training Time ◽  
Training Support ◽  
Large Scale Data ◽  
Vector Machines ◽  
Speech Classification ◽  
Scale Data

SVM has got very good results in the area of solving the classification, regression and density estimation problem in machine learning, has been successfully applied to practical problems of text recognition, speech classification, but the training time is too long is a big drawback. A new reduction strategy is proposed for training support vector machines. This method is fast in convergence without learning machine’s generalization performance, the results of simulation experiments show the feasibility and effectiveness of that method through this method.

Large-scale pinball twin support vector machines

2021 ◽  
Author(s):  
M. Tanveer ◽  
A. Tiwari ◽  
R. Choudhary ◽  
M. A. Ganaie
Keyword(s):  
Support Vector Machines ◽  
Large Scale ◽  
Support Vector ◽  
Twin Support Vector Machines ◽  
Vector Machines

scholarly journals An iterative method for classification of binary data

2020 ◽  
Author(s):  
Denali Molitor ◽  
Deanna Needell
Keyword(s):  
Binary Data ◽  
Large Scale ◽  
Support Vector ◽  
Large Scale Data ◽  
Classification Framework ◽  
Vector Machines ◽  
Inference Methods ◽  
Compressed Data ◽  
Scale Data

Abstract In today’s data-driven world, storing, processing and gleaning insights from large-scale data are major challenges. Data compression is often required in order to store large amounts of high-dimensional data, and thus, efficient inference methods for analyzing compressed data are necessary. Building on a recently designed simple framework for classification using binary data, we demonstrate that one can improve classification accuracy of this approach through iterative applications whose output serves as input to the next application. As a side consequence, we show that the original framework can be used as a data preprocessing step to improve the performance of other methods, such as support vector machines. For several simple settings, we showcase the ability to obtain theoretical guarantees for the accuracy of the iterative classification method. The simplicity of the underlying classification framework makes it amenable to theoretical analysis.

Using PROSEPCT and SVM for the Estimation of Chlorophyll Concentration

2014 ◽  
Vol 989-994 ◽  
pp. 2184-2187
Author(s):  
Jie Lv ◽  
Feng Li Deng ◽  
Zhen Guo Yan
Keyword(s):  
Chlorophyll Content ◽  
Large Scale ◽  
Chlorophyll Concentration ◽  
Support Vector ◽  
Maximum Correlation ◽  
Dynamic Prediction ◽  
West Lake ◽  
Chlorophyll Contents ◽  
Irrigated Area ◽  
Vector Machines

This study focused on estimating chlorophyll concentration of rice using PROSPECT and support vector machine. The study site is located in West Lake sewage irrigation area of Changchun, Jiliin Province. Reflectance spectrual of rice were measured by ASD3 spectrometer, chlorophyll contents of rice were recorded with a portable chlorophyll meter SPAD-502. Support vector machines and PROSPECT model were adopted to construct hyperspectral models for predicting chlorophyll content. The results indicate that: the hyperspectral prediction model of rice chlorophyll content yields a maximum correlation coefficient of 0.8563, and achieves a smallest RMSE of 9.5106; and the prediction accuracy based on the first derivative spectrum is higher than on the original spectrum. Research of this paper provides a theoretical basis for large scale dynamic prediction of rice chlorophyll content in sewage irrigated area.

Sign in / Sign up

Export Citation Format

Share Document