Regularization methods

Several numerical procedures for EADF evaluation have also been proposed. Morrison and Ross [19] developed the so-called CAEDMON (Computed Adsorption Energy Distribution in the Monolayer) method. Adamson and Ling [20] proposed an iterative approximation that needs no a priori assumptions. Later, House and Jaycock [21] improved that method and proposed the so-called HILDA (Heterogeneity Investigation at Loughborough by a Distribution Analysis) algorithm. Stanley et al. [22,23] presented two regularization methods as well as the method of expectation maximalization. 

In this chapter we study the stability with respect to the initial data and the right-hand side of two-layer and three-layer difference schemes that are treated as operator-difference schemes with operators in Hilbert space. Necessary and sufficient stability conditions are discovered and then the corresponding a priori estimates are obtained through such an analysis by means of the energy inequality method. A regularization method for the further development of various difference schemes of a desired quality (in accuracy and economy) in the class of stability schemes is well-established. Numerous concrete schemes for equations of parabolic and hyperbolic types are available as possible applications, bring out the indisputable merit of these methods and unveil their potential. 

Constructions of economical factorized schemes. Using the regularization method behind, we try to develop the general method for constructing stable economical difference schemes on the basis of the primary stable scheme... 

Morozov, V. (1984) Regularization Methods for Solving Improperly Posed Problems. Springer New-York-Berlin-Heidelberg. 

K. G. Hollingsworth, M. L. Johns 2003, (Measurement of emulsion droplet sizes using PFG NMR and regularization methods),/. Colloid Interface Sci. 258, 383. 

Several formylfuranboronic acids and similar compounds have been prepared by regular methods and studied with respect to their 13C and H NMR spectra. As in the benzene series, the protons ortho to the boron are deshielded and the meta protons are shielded. Good conjugation exists between the furan ring (donor) and both the boron atom and the carbonyl group (acceptors). A Hammett-Jaffe analysis yielded ap+ + 0.256 for the B(OH)2 group.234... 

The import of diabatic electronic states for dynamical treatments of conical intersecting BO potential energy surfaces is well acknowledged. This intersection is characterized by the non-existence of symmetry element determining its location in nuclear space [25]. This problem is absent in the GED approach. Because the symmetries of the cis and trans conformer are irreducible to each other, a regularization method without a correct reaction coordinate does not make sense. The slope at the (conic) intersection is well defined in the GED scheme. Observe, however, that for closed shell structures, the direct coupling of both states is zero. A configuration interaction is necessary to obtain an appropriate description in other words, correlation states such as diradical ones and the full excited BB state in the AA local minimum cannot be left out the scheme. 

In general, problems having solutions that vary radically or discon-tinuously for small input changes are said to be ill-posed. Deconvolution is an example of such a problem. Tikhonov was one of the earliest workers to deal with ill-posed problems in a mathematically precise way. He developed the approach of regularization (Tikhonov, 1963 Tikhonov and Arsenin, 1977) that has been applied to deconvolution by a number of workers. See, for example, papers by Abbiss et al (1983), Chambless and Broadway (1981), Nashed (1981), and Bertero et al. (1978). Some of the methods that we have previously described fall within the context of regularization (e.g., the method of Phillips and Twomey, discussed in Section V of Chapter 3). Amplitude bounds, such as positivity, are frequently used as key elements of regularization methods. 

You may notice that the ridge regression is a straightforward statistical counterpart of the regularization methods discussed in Section 1.7. 

Determination of volatile matter content using a slower heating rate is applicable to a wider variety of coals. However, the values obtained are sometimes lower (1 to 3% absolute) than those obtained from the regular method. This illustrates the empirical nature of this test and the importance of strict adherence to detailed specifications. The complexity of the constituents of coal that undergo decomposition during this test makes it necessary to have wide tolerances for reproducibility and repeatability. 

S.W. Provencher, A constrained regularization method for inverting data represented by linear algebraic or integral equations, Comput. Phys. Commun. 27 (1982) 213-227. 

Although the methods previously discussed have shown some promise in solving the Fredholm equation, two recent methods were chosen for intensive study and comparison with our GEX function fit method. These methods were the constrained regularization method, and the polynomial subdistribution method. 

For both the subdistribution and the GEX fit methods a Marquardt algorithm for constrained non-linear regression was used to minimize the sum of squares error (.10). The FORTRAN program CONTIN was used for the constrained regularization method. All computations were performed on a Harris H-800 super mini computer. 

Besides Tikhonov regularization, there are numerous other regularization methods with properties appropriate to distinct problems [42, 53,73], For example, an iterated form of Tikhonov regularization was proposed in 1955 [77], Other situations include using different norms instead of the Euclidean norm in Equation 5.25 to obtain variable-selected models [53, 79, 80] and different basis sets such as wavelets [81],... 

Tikhonov, A.N., Solution of incorrectly formulated problems and the regularization method, Soviet Math. Dokl., 4, 1035-1038, 1963. 

As for the non-uniqueness of the solution, there is no method that can bypass this inherent problem. In inverse problems, one of the common practices to overcome the stability and non-uniqueness criteria is to make assumptions about the nature of the unknown function so that the finite amount of data in observations is sufficient to determine that function. This can be achieved by converting the ill-posed problem to a properly posed one by stabilization or regularization methods. In the case of groundwater pollution source identification, most of the time we have additional information such as potential release sites and chemical fingerprints of the plume that can help us in the task at hand. 

The integral equation (145) presents a classic example of an ill-posed problem, by which one means that the solution i/dx) does not depend continuously on the data function R(X). In the above formulation of the problem, R(X) is known only for X Xj (j = 1,2,..., m) and the data are given with known errors AR/Xj). With these inadequate data, it is extremely difficult, in general, to solve Eq. (145) (see e.g. Ref. 329). One possible approach is to apply the statistical regularization method (STREG) [330]. 

To solve this problem we use the simplex regular method. For k= 5, the dimensionless matrix of experiments is obtained with relation (5.138). Thus, the matrix of the dimensionless factors is transformed into dimensional values with relations (5.96) and (5.97). Table 5.33 corresponds to this matrix, the last column of which contains the values of the process response. According to this table, the point placed in position 4 was found to be the least favourable for the process. However, before rejecting it, we have to build the coordinates of the new point by means of the image reflection of point number 4 (this point will be calculated to be number 7 from k-i-l-i-1). For this purpose, we use relations (5.145) and (5.146). 

Honerkamp J and Weese J (1989) Determination of the relaxation spectrum by a regularization method. Macromolecules 22 4372-7. 

J.Honerkamp, J.Weese, A non linear regularization method for the calculation of relaxation spectra, Rheol. Acta 22 (1993), 65-73. 

The methods depend on the theoretical treatment which is used. A majority of them are based on the Generalised Adsorption Isotherm (GAI) also called the Integral Adsorption Equation (LAE). The more recent approaches use the Monte Carlo simulations or the density functional theory to calculate the local adsorption isotherm. The analytical form of the pore size distribution function (PSD) is not a priori assumed. It is determined using the regularization method [1,2,3]. Older methods use the Dubinin-Radushkevich or the Dubinin-Astakhov models as kernel with a gaussian or a gamma-type function for the pore size distribution. In some cases, the generalised adsorption equation can be solved analytically and the parameters of the PSD appear directly in the isotherm equation [4,5,6]. Other methods which do not rely on the GAI concept are sometimes used the MP and the Horvath-Kawazoe (H-K) methods are the most well known [7,8]. 

Alvares et. al. [141] successfully applied a method known as the Tikhonov Regularization method and L-curve criterion to generate data in close accord with the Malvern software. 

---