Support Vector Machine Optimized by Fireworks Algorithm for Handwritten Digit Recognition

摘要: Handwritten digit recognition is an important subarea in the object research area. Support vector machines represent a very successful recent binary classifier. Basic support have to be improved order deal with real-world problems. The introduction of soft margin for outliers and misclassified samples as well kernel function non linearly separably data leads hard optimization problem selecting parameters these two modifications. Grid search which often used rather inefficient. In this paper we propose use one latest swarm intelligence algorithms, fireworks algorithm, machine tuning. We tested our approach on standard MNIST base handwritten images selected set simple features obtained better results compared other approaches from literature.