Predictions using aggregate information

摘要

Predictions of a given predictive model may be improved using aggregate information. A plurality of targets to predict in a given domain may be identified, and may be predicted based on raw data set. Aggregate information associated with the plurality of targets is received, the aggregate information including estimated or actual values at a coarser level of the plurality of targets, and based on the aggregate information, the predicted target values may be improved in prediction accuracy.

主权项

1. A method for improving predictions of a given predictive model using aggregate information, comprising:
identifying a plurality of targets to predict in a given domain; receiving raw data set for predicting a plurality of values associated respectively with the plurality of targets in the given domain and predicting the plurality of values; receiving aggregate information associated with the plurality of targets, the aggregate information including estimated or actual values at a coarser level of the plurality of targets; and improving, by a processor, the predicted values of the plurality of targets based on the aggregate information, wherein the improving comprises: summing the plurality of values to determine an aggregated value; subtracting the aggregated value from the aggregate information to determining a corresponding difference; dividing the corresponding difference by a count of the plurality of targets to determine a factor; and adding the factor to each of the plurality of values, wherein the plurality of values with the added factor represents the improved predicted values of the plurality of targets.