Given a set of objects and a training data set of interactions between some pairs of those objects (not exhaustively all pairs) with each such interaction labeled positive or negative, learn a model which predicts the outcome of pairs not seen in the training data.
Inspiration is online dating or hookups.
Assuming a graphical model, what interesting hidden variables will be learned? Lovemaking style (typically a feature difficult to quantify or objectively measure)?
No comments :
Post a Comment