Month: April 2019

These days I’m taking a course for graduate students called numerical method. The concepts and ideas are too tough for me, and I think its a good idea to write down something about the methods, and to make a comparison … Read More

Let’s say you have input data x and you want to classify the data into labels y. A generative model learns the joint probability distribution p(x,y) and a discriminative model learns the conditionalprobability distribution p(y|x) – which you should read as “the probability of y given x“. Here’s a really simple example. Suppose you … Read More