Kernel method has to do with machine learning. Kernel method is a class of algorithm which is used for pattern analysis. A common example of kernel method is support vector machine. Advantages of support vector machine over deep learning include the following: it is less prone to over drifting, it produces sparse solutions, it does not limit the number of features and it automatically select the model size. Deep learning refers to applications to learning tasks of artificial nerural network that have more than one hidden layer.
