基于DESTIN-SVM的模式识别

The Pattern Recognition Based on DESTIN-SVM

提出了基于DESTIN-SVM模型的模式识别方法,给出了该模型的训练和识别方法。DESTIN模型一般采用神经网络作为分类器,训练时间长,使用SVM取代神经网络集成到DESTIN模型中,增强了系统的分类识别能力,并且用CUDA对算法进行加速,提高了系统的识别速度。

A pattern recognition method based on DESTIN-SVM model is proposed,and the training and recognition method of the model is given.The DESTIN model generally uses neural network as classifier,and the training time is long.By using SVM to replace the neural network in the DESTIN model,the recognition ability of the system is enhanced,and the algorithm is accelerated by using CUDA,which improves the recognition speed of the system.