A multiresolution approach for the convergence acceleration of multivariate curve resolution methods

摘要: Modern computerized spectroscopic instrumentation can result in high volumes of data. Such accurate measurements rise special computational challenges for multivariate curve resolution techniques since pure component factorizations are often solved via constrained minimization problems. The costs these calculations rapidly grow with an increased time or frequency the spectral measurements. key idea this paper is to define given high-dimensional data a sequence coarsened subproblems reduced resolutions. multiresolution algorithm first computes factorization coarsest problem lowest resolution. Then results used as initial values next higher Good fast solution on refined level. This procedure repeated and finally determined highest level described approach allows considerable convergence acceleration. analyzed tested experimental from rhodium-catalyzed hydroformylation together various soft hard models.