EnsembleSVM: A Library for Ensemble Learning Using Support Vector Machines
This work provides a software tool for researchers and practitioners in machine learning to efficiently implement ensemble methods with SVMs, though it is incremental as it builds on existing ensemble and SVM techniques.
The authors tackled the problem of high training complexity in ensemble learning with support vector machines by developing the EnsembleSVM library, which avoids duplicate storage and evaluation of shared support vectors, resulting in drastically reduced training complexity while maintaining high predictive accuracy.
EnsembleSVM is a free software package containing efficient routines to perform ensemble learning with support vector machine (SVM) base models. It currently offers ensemble methods based on binary SVM models. Our implementation avoids duplicate storage and evaluation of support vectors which are shared between constituent models. Experimental results show that using ensemble approaches can drastically reduce training complexity while maintaining high predictive accuracy. The EnsembleSVM software package is freely available online at http://esat.kuleuven.be/stadius/ensemblesvm.