Using Multiple Models to Understand Data

Proceedings of the International Joint Conference on Artificial Intelligence (IJCAI 2011) |

Published by AAAI Press

Publication

A human’s ability to diagnose errors, gather data, and generate features in order to build better models is largely untapped. We hypothesize that analyzing results from multiple models can help people diagnose errors by understanding relationships among data, features, and algorithms. These relationships might otherwise be masked by the bias inherent to any individual model. We demonstrate this approach in our Prospect system, show how multiple models can be used to detect label noise and aid in generating new features, and validate our methods in a pair of experiments.