I have learnt about regression (linear regression, logistic regression and others) and also heard about classification.
What is the difference?
To me it seems that those are very similar problems.
Thanks in advance.
That is a good question that I think many persons have when studying data mining.
When we talk about regression, in general we want to predict a value that is numerical and often this will be done based on numerical data. For example, we can try to fit a linear equation to some numerical data points from a time series to try to predict the next data point which is a number.
In classification, we generally aim to predict an attribute value that is not numeric. Typically, we want to predict class or group that some objects might belong to. For example, given the description of a bank customer, we may want to predict if he belongs to customers who will pay back their loan or the customers who will not pay back.
Hope that this gives you some ideas about this. This is just a brief explanation.