Hello. As a newbie to machine learning problem that fits these facts about my data:
1) My inputs are all numbers and my output is a number. 2) I have many features/inputs of varying importance and varying interrelatedness. 3) I have basically infinite training data (sets of both input and the correct output) 4) A linear regression would be absurdly inaccurate for almost all cases because of how little the true model looks like a line or even a single continuous function. 5) I only need to get the formula once.
Are any machine learning approaches suited for these cases out of the box?
[link] [7 comments]