Neuronal morphology classification based on SVM

Neuron classification is the research basis and also a difficult issue for neuroscience. In this paper, a novel neuronal morphology classification method based on support vector machine (SVM) was proposed. In this method, we first estimated the neuronal geometrical morphological features according to the original space geometric data. Then we utilized SVM to classify the neurons based on the new morphological features. Essentially, this method converts the neuronal morphology classification problem to a quadratic optimization problem using non-linear transformation and structural risk minimization, which performs high accuracy and stability. Experimental results show that the proposed method is effective.