Dimensionality Reduction of Reactor Noise Signatures

A minimum-error linear transformation was applied to reduce the dimensionality of reactor noise signatures. The procedure improves the computational efficiency of processing algorithms and decreases bulk data storage requirements. The method was tested with noise signatures from the High-Flux Isotope Reactor at the Oak Ridge National Laboratory, and a dimensionality reduction in excess of 90% was achieved without loss in the average characteristics of the signatures.