关于Query by Committee (QBC)的描述

Case 1:

The QBC framework quantifies the prediction uncertainty based on the level of disagreement among an ensemble of matrix completion algorithms. Specifically, a committee of matrix completion algorithms are applied on the partially observed data matrix to impute the missing values. The variance of prediction (among the committee members) of each missing entry is taken as a measure of uncertainty of that entry. The top k uncertain entries are then queried for manual annotation.

 

1. The most informative query, then, is the instance over which the committee is in most disagreement about how to label.