A new method is introduced for combining information from multiple sources to support one-class classification. The contributing sources may represent measurements taken by different sensors of the same physical entity, repeated measurements by a single sensor, or numerous features computed from a single measured image or signal. The approach utilizes the theory of statistical hypothesis testing, and applies Fisher's technique for combining p-values, modified to handle nonindependent sources. Classifier outputs take the form of fused p-values, which may be used to gauge the consistency of unknown entities with one or more class hypotheses. The approach enables rigorous assessment of classification uncertainties, and allows for traceability of classifier decisions back to the constituent sources, both of which are important for high-consequence decision support. Application of the technique is illustrated in two challenge problems, one for skin segmentation and the other for terrain labeling. The method is seen to be particularly effective for relatively small training samples.
X-ray fluorescence (XRF) has been employed as one of several orthogonal means of screening materials to prevent counterfeit and adulterated products from entering the product stream. We document the use of principal component analysis (PCA) of XRF data on compositionally similar and dissimilar stainless steels for the purpose of testing the feasibility of employing XRF spectra to parse and bin these alloys as the same or significantly different alloy materials. The results indicate that XRF spectra can separate and assign alloys via PCA, but that important corrections for detector drift and scaling must be performed in order to achieve valid results.
Mangolini, Filippo M.; Mangolini, Filippo M.; Koshigan, Komlavi D.; Koshigan, Komlavi D.; Van Benthem, Mark V.; Van Benthem, Mark V.; Ohlhausen, J.A.; Ohlhausen, J.A.; McClimon, J B.; McClimon, J B.; Hilbert, James A.; Hilbert, James A.; Fontaine, Julien F.; Fontaine, Julien F.; Carpick, Robert W.; Carpick, Robert W.