Ask HN: How can I write a simple ML algorithm

Hi HN,

I want to write a simple ML algorithm in python for picking out good objects out of a large set of objects.

Say each object has feature(X1,X2,....XN) which are all numeric for the sake of simplicity.Now I have a billion such objects and a training set of good objects.Now based on the training set how do I pick out good objects from the large set.

One obvious solution is to take the mean of every feature in the training set and then find out objects whose values are close to the mean or more precisely the whose sum of deviations(or should it be sum of squares of deviations?) is the lowest.But I am looking for something cooler with ML?.Can anyone point me in the right direction?

Thanks