Deep Support Vector Machines for Regression Problems

One of the most accurate machine learning algorithms nowadays is the Support Vector machine. Support Vector Machines use kernels in order to project data on the featurespace. We will introduce a new method of machine learning dubbed the Deep Support Vector Machine. Instead of using a kernel, the Deep Support Vector Machine tries to extract features from its input in order to project it on the featurespace. Now instead of using predefined kernels to classify data we are able to make classifications based on the features of a given input vector.