Prior information

By imposing the prior information on the solution one implies, that this solution is suited to the only class of problems which corresponds to the information that is involved. But it should be noticed that these classes can be so wide that real constraints to the restored image are reduced to minimum. 

Finally, it is shown in terms of the presented example that the proposed adaptive reconstruction algorithm is valuable for image reconstruction from projections without any prior information even in the case of noisy data. The number of required projections can be determined by investigating the convergence properties of the reconstruction algorithm. 

All three tasks are generally too complicated to be solved from first principles. They are, therefore, tackled by making use of prior information, and of information that has been condensed into knowledge. The amount of information that has to be processed is often quite large. At present, more than 41 million different compounds are known all have a series of properties, physical, chemical, or biological all can be made in many different ways, by a wide range of reactions all can be characterized by a host of spectra. This immense amount of information can be processed only by electronic means, by the power of the computer. 

International Code of Conduct on the Distribution and Use of Pesticides, United Nations Pood and Agriculture Organization, New York, 1985 amended to include provisions for "Prior Informed Consent," 1990. 

Similarly, prior information distributed as a gamma lunctiou with an exponentially distributed update gives a posterior that also is gamma distributed. 

It may be noted that if 0= 1 and r= 0, the prior, equation 2.6-11, becomesp(/I) = 1. This is the flat prior indicating no prior information that leads to classical results when these parameters are inserted into equation 2.6-12. 

Truncating the plane constrains the centroid estimate to a certain region, making the variance finite. Since the truncated plane is placed where the centre is expected to be we are implicitly adding prior information (van Dam and Lane, 2000). The smaller the plane, the more the centroid is effectively localized and the more prior information is assumed. Therefore, by adding prior information, truncating the plane can improve the centroid estimate, even though some photons are lost. The optimal solution is to maximize the likelihood directly. 

Any prior information that is available about the parameter values can facilitate the estimation of the parameter values. The methodology to incorporate such information in the estimation procedure and the resulting benefits are discussed in Chapter 8. 

Under certain conditions we may have some prior information about the parameter values. This information is often summarized by assuming that each parameter is distributed normally with a given mean and a small or large variance depending on how trustworthy our prior estimate is. The Bayesian objective function, SB(k), that should be minimized for algebraic equation models is... 

We have assumed that the prior information can be described by the multivariate normal distribution, i.e., k is normally distributed with mean kB and co-variance matrix VB. 

Practical experience has shown that (i) if we have a relatively large number of data points, the prior has an insignificant effect on the parameter estimates (ii) if the parameter estimation problem is ill-posed, use of "prior" information has a stabilizing effect. As seen from Equation 8.48, all the eigenvalues of matrix A are increased by the addition of positive terms in its diagonal. It acts almost like Mar-quadt s modification as far as convergence characteristics are concerned. 

Let us illustrate the methodology by consider cases (i) and (ii). We assume that from a preliminary experiment or from some other prior information we have the following parameter estimates for each model ... 

In practice when reservoir parameters such as porosities and permeabilities are estimated by matching reservoir model calculated values to field data, one has some prior information about the parameter values. For example, porosity and permeability values may be available from core data analysis and well test analysis. In addiction, the parameter values are known to be within certain bounds for a particular area. All this information can be incorporated in the estimation method of the simulator by introducing prior parameter distributions and by imposing constraints on the parameters (Tan and Kalogerakis, 1993). 

It is reasonable to assume that the most probable values of the parameters have normal distributions with means equal to the values that were obtained from well test and core data analyses. These are the prior estimates. Each one of these most probable parameter values (kBj, j=l,...,p) also has a corresponding standard deviation parameter estimate. As already discussed in Chapter 8 (Section 8.5) using maximum likelihood arguments the prior information is introduced by augmenting the LS objective function to include... 

While prior information may be used to influence the parameter estimates towards realistic values, there is no guarantee that the final estimates will not reach extreme values particularly when the postulated grid cell model is incorrect and there is a large amount of data available. A simple way to impose inequality constraints on the parameters is through the incorporation of a penalty function as already discussed in Chapter 9 (Section 9.2.1.2). By this approach extra terms are added in the objective function that tend to explode when the parameters approach near the boundary and become negligible when the parameters are far. One can easily construct such penalty functions. For example a simple and yet very effective penalty function that keeps the parameters in the interval (kmjnkmaXil) is... 

If one wishes to use NO PRIOR INFORMATION on a particular parameter simply use Vprior = 0 (i.e., variance is infinity - no prior info). 

It appears from formula (6) that the prior-prejudice distribution mix) is a fundamental quantity in the calculation of the MaxEnt distribution of electrons, in that the latter is obtained by modulation of m(x). In all those regions where the modulating factor required to fit the observations is unity, the final picture is therefore always going to coincide with the prior expectation itself. For this reason, it is of the greatest importance that some of the prior information available about the system under study be conveyed into the calculation by means of a sensible choice for the prior-prejudice distribution. 

Generally, extensive prior information is known about the sample in terms of the elemental composition and in these cases methods of analysis can be selected that will provide the desired result. However, if this information is not available or else a more general survey of ultratrace elements is required, then AES with an inductively coupled plasma source is the only atomic spectrometric technique that can provide these data at ultratrace levels. 

Group comparison tests for proportions notoriously lack power. Trend tests, because of their use of prior information (dose levels) are much more powerful. Also, it is generally believed that the nature of true carcinogenicity (or toxicity for that matter), manifests itself as dose-response. Because of the above facts, evaluation of trend takes precedence over group comparisons. In order to achieve optimal test statistics, many people use ordinal dose levels (0,1,2..., etc.) instead of the true arithmetic dose levels to test for trend. However, such a decision should be made a priori. The following example demonstrates the weakness of homogeneity tests. 

Model building remains a useful technique for situations where the data are not amenable to solution in any other way, and for which existing related crystal structures can be used as a starting point. This usually happens because of a combination of structural complexity and poor data quality. For recent examples of this in the structure solution of polymethylene chains see Dorset [21] and [22]. It is interesting to note that model building methods for which there is no prior information are usually unsuccessful because the data are too insensitive to the atomic coordinates. This means that the recent advances in structure solution from powder diffraction data (David et al. [23]) in which a model is translated and rotated in a unit cell and in which the torsional degrees of freedom are also sampled by rotating around bonds which are torsionally free will be difficult to apply to structure solution with electron data. 

The use of constraints and restraints arise from unrealistic bond lengths and angles. It is a reasonable use of statistical analysis to include prior information in a refinement procedure, and geometry is one such restraint/constraint. Poor geometries are often a consequence of missing data down one axis, invariably the result of the missing cone or lack of an epitaxially grown crystal. This lack of data has a profound effect on the refinement process... 

A thorough study by Rosenberg et al. (4) examined human population structure using 377 markers in 1056 individuals from 52 populations around the world. Without prior information about the origins of individuals, these authors used a Bayesian algorithm to identify six major genetic clusters (1) sub-Saharan Africans ... 

In this as in all other problems of experimental design, prior information is helpful. For example, if the enzyme we are dealing with is naturally found in a neutral environment, then it would probably be most active at a neutral pH, somewhere near pH = 7. If it were found in an acidic environment, say in the stomach, it would be expected to exhibit its optimal activity at a low (acidic) pH. When information such as this is available, it is appropriate to center the experimental design about the best guess of where the desired region might be. In the absence of prior information, factor combinations might be centered about the midpoint of the factor domain. 

---