Peter Binev

We consider the problem of learning an unknown function f from given data about f. The learning problem is to give an approximation fˆ to f that predicts the values of f away from the data. There are numerous settings for this learning problem...