No difference unsupervised learning and supervised learning

Supervised learning: given a training set, then set out to learn by training a model, when new data comes, we will be able to predict the results of this model. Supervised learning the training set not only have input but also output, ie, the target characteristics and objective results, data unique feature is the training set, the result is a marked man. Supervised learning is our beginning and let the machine know a picture is something that, as he is a pig, he is a dog, he is a cow, the machine will get to determine a set of parameters by learning that he is a pig, a dog, or cow, when when we enter a picture again, he will be the model parameters to determine than what he is

Unsupervised Learning: Enter thing is not marked in advance, nor the result of determination, we enter a bunch of photos, the photo we have cattle, pigs and dogs, but we did not tell the machine in advance, what he is, the machine will be learn to distinguish pigs, cattle and dogs, but the machine does not know he is a pig, cow or dog, because we did not tell the machine what we gave him photographs, machine distinction is classified by the similarity of the sample set.