Streamline how "interestingness image" is computed

[russelldj]

Many planners take in an interestingness image. This is currently most often the uncertainty of an uncertain predictor. But it could be other things, such as the predicted likelihood of a rare class. It would be nice to have a set of functions that take in the predictions from a predictor and return an interestingness image.