A fast non-negativity-constrained least squares algorithm

Abstract

In this paper a modification of the standard algorithm for non‐negativity‐constrained linear least squares regression is proposed. The algorithm is specifically designed for use in multiway decomposition methods such as PARAFAC and N‐mode principal component analysis. In those methods the typical situation is that there is a high ratio between the numbers of objects and variables in the regression problems solved. Furthermore, very similar regression problems are solved many times during the iterative procedures used. The algorithm proposed is based on the de facto standard algorithm NNLS by Lawson and Hanson, but modified to take advantage of the special characteristics of iterative algorithms involving repeated use of non‐negativity constraints. The principle behind the NNLS algorithm is described in detail and a comparison is made between this standard algorithm and the new algorithm called FNNLS (fast NNLS). © 1997 John Wiley & Sons, Ltd.

Properly handling negative values in the calculation of binding constants by physicochemical modeling of spectroscopic titration data

“…A variety of algorithms achieving this nonnegative regression are readily available, including Matlab's built-in lsqnonneg algorithm and the fast nonnegative least squares (FNNLS) algorithm published by Bro and De Jong. 47 We find the latter most efficient while giving equivalent results. The third approach is to simply perform unconstrained regression and accept the negative results.…”

High-Entropy Alloys: Assessing Atomic-Scale Mixing and Surface Passivation with Time-of-Flight Secondary Ion Mass Spectrometry

“…This variation of the energy per atom (depending on the nature of primary ion) influences the ion penetration in the solid and the sputtering process, including the secondary ion yields and fragmentation patterns observed in ToF-SIMS. 40,44 To best quantify the atomic degree of mixing of the samples, the high mass secondary clusters are the most important since they can be taken as the markers of the atomic mixing in the alloy. Similarly, for examination of the oxidation resistance of the samples, the secondary clusters of interest are oxides.…”

Multiscale and Multimodal Image Fusion. Coping with Differences in Scanned Area and Spatial Resolution for Raman/Fluorescence Images of Labeled Cells

“…16 Non-negativity was applied to the concentration and spectral profiles of both multisets since fluorescence and Raman spectra cannot have negative values. 17 In the Raman analysis, the correspondence of species constraint, linked to the definition of presence or absence of some constituents in the images analyzed, was used since the scanned area of some of the images did not include certain cellular components. 18 Table 2 summarizes the main MCR-ALS results obtained in all multisets analyzed.…”