Classification Problems in Machine Learning

Classification:

As have been discussed earlier, a supervised problem is a classification problem if the output it produces is discretized.

Some examples of classification problems are classifying if an email is spam or not. Are online transactions fraudulent or not, a tumor is malignant or benign.

What do we actually do to predict outputs?

First, as we have told earlier, a hypothesis H is formed using the training data. Then, whenever we want to predict the output for a set of features, we feed it to H. In case, the output comes to be greater than 0.5 we predict 1 else we predict 0.