Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for analyzing outcome estimation models. In one aspect, a method includes providing tracer data to a content performance estimation model. The tracer data specify feature values for a reference content item that is not distributed to users, and for which a reference performance measure is pre-defined. An estimated performance measure that was computed for the reference content item is received from the performance estimation model. An accuracy measure is computed for the content performance estimation model using the estimated performance measure and the reference performance measure. Data are provided that specify the accuracy measure for the content performance estimation model.

Abstract: Disclosed are a method and system for receiving a plurality of potential features to be added to a model having existing features. For each of the potential features, an approximate model is learned by holding values of the existing features in the model constant. The approximate model includes the model having existing features and at least the potential feature. A performance metric is computed for evaluating performance of the approximate model. The performance metric is used to rank the potential feature based on a predetermined criterion.

Abstract: Methods of operating a robot comprised of a plurality of modules during reconfiguration, rearrangement to provide a support structure and internal manipulation, each module made of a cubic shape with six faces, each face including an arm for expansion and contraction with another module.

Abstract: The invention provides methods of operating a robot comprised of a plurality of modules during reconfiguration, rearrangement to provide a support structure and internal manipulation, each module made of a cubic shape with six faces, each face including an arm for expansion and contraction with another module.