Optimized SVM using AdaBoost and PSO to Classify Brain Images of MR

In current paper, it is suggested a technique for improving a model of feature detection based on AdaBoost, weighted support vector machine (WSVM) using particle swarm optimization (PSO) for selection of features and diagnose of Alzheimer disease (AD) classification problems. Our paper contributions can be stated as it was for the first time employing this method aimed at classification of brain magnetic resonance(MR) imaging with very good classification accuracy. Moreover, our suggested scheme is quite appropriate in dealing through high amount of data (sparse data) to classy of the brain image. The data used in this study consisted of 198 Alzheimer disease (AD) data and 229 normal control (NC), which used for learning and testing. The final results of this study displays that proposed method classification accuracy is 93% which is promising performance.