METHODS AND SYSTEMS TO EXTRACT SIGNALS FROM LARGE AND IMPERFECT DATASETS

A variety of signal extraction algorithms are employed to construct indicators or synthetic data proxy variables or factors from one or multiple data sources with specialized data handling options to resolve problems associated with economic and financial data: revisions, missing observations, publication lags, spontaneous arrival. A forecast evaluation module aids in the selection of the signal extraction procedure that has produced the most accurate factor.