Data interpretation

At each stage of a field life cycle raw data has to be converted into information, but for the information to have value it must influence decision making and profitability. 

The relationship for the interpretation of data measured for RTD can be easily derived with Laplace transformation [54]. However, the equations so obtained are not convenient for time-dispersed measurement because they still need Laplace transformation to deal with the data. In the following, another relationship will be derived. 

Consider the equipment system schemed in Fig. 3.8, where the subscript B denotes the process particles and A the tracer. Corresponding to the impinging stream equipment with feeding on both sides, the system has two feeds, which are denoted by the superscripts (1) and (2), respectively while it has only one out stream of particles. For convenience of operation, the tracer A is inputted into one feeding stream, i.e., (2)  

At the instant r, the residence times of all the particles of the tracer A must be f while the particles of B consist of two groups in the first group all the particles are fed at instant f0 = 0 or later and so also have residence times t while the particles in the second group were fed into the device before tQ = 0 and so have residence times t. If the particles in the first group are denoted by the superscript then, from the definition of the residence time distribution function, F, we have 

On the other hand, one of the essential conditions for correct measurement of RTD is that the tracer particles have the properties, including RTD characteristics, very close to those of the process particles. This implies that, in the time domain of t 0, the residence time distribution functions of particles A and B in the same device should be approximately equal to each other, i.e. 

It should be noted that Eq. (3.31) is an approximate relationship. This is because both the individual flow rates of the particles A and B vary with time, although the total flow rate, i.e. the flow rate of A plus that of B, is stable, while the variation of particle flow rate may affect RTD. However, since the variations of the flow rates of particles A and B are not so large, the possible deviation of Function F caused by this factor may be neglected, just like the generalized (dimensionless) residence time distribution function can be extended for application in a certain range [57]. 

This part is probably the most difficult to manage when it comes to precision agronomy. Farmers can be overwhelmed by the amount of data being produced, and the numerous interactions that might be taking plaee in the field. There are some fimdamental questions to be asked when looking at data from a field or a erop  

When the precision fanning systems were in their infancy, it was often difficult to interpret yield maps, for instance when they were all the farmer had to go on. One cormnon dilermna was If I have an area of low yield in a field, should I put more fertihser on it to try and bring the yield up, or accept that it will always be low yielding because of other factors, and I need to put less on, thus saving money on wasted material  

In earlier chapters, the diagnostic values, organ and tissue specificity, and limitations of some tests have been discussed, and this knowledge should be applied when interpreting data. A test value change due to cellular injury should be distinguished 

The findings of biochemical effects in preclinical toxicology studies are relatively common Heywood (1981, 1983) found more than 30% of a mixture of tested compounds affected blood chemistry measurements in rats, dogs, and monkeys there were species differences, and these findings were less common in monkeys. In 61 subchronic rat studies conducted by the U.S. National Toxicology Program, liver and kidney lesions were reported in 31 and 41% of the studies, respectively (Travlos et al. 

but the incidence of these findings will depend on the discovery and development interests of the particular organization. 

Anderson, N.L., Anderson, N.G. (2001). The human plasma proteome history, character, and diagnostic prospectss. Mol. Cell. Proteomics 1, 845-867. 

Apffel, A. (2004). Multidimensional chromatography of intact proteins. In Simpson, R.J., editor, Purifying Proteins for Proteomics A Laboratory Manual. Cold Spring Harbor Laboratory Press Cold Spring Harbor, New York pp.75-100. 

Apffel, A., Fischer, S., Goldberg, G., Goodley, P.C., Kuhlmann, F.E. (1995). Enhanced sensitivity for peptide mapping with electrospray liquid chromatography—mass spectrometry in the presence of signal suppression due to trifluoroacetic acid-containing mobile phases. J. Chromatogr. A 712, 177-190. 

Benesch, J.L., Robinson, C.V. (2006). Mass spectrometry of macromolecular assemblies preservation and dissociation. Curr. Opin. Struct. Biol. 6, 245-251. 

Brunet, S., Thibault, P., Gagnon, E., Kearney, P, Bergeron, J.J., Desjardins, M. (2003). Organelle proteomics looking at less to see more. Trends. Cell. Biol. 13, 629-638. 

The cyclic voltammogram is characterized by several important parameters. Four of these observables, the two peak currents and two peak potentials, provide the basis for the diagnostics developed by Nicholson and Shain (1) for analyzing the cyclic voltammetric response. 

1 Reversible Systems The peak current for a reversible couple (at 25°C), is given by the Randles-Sevcik equation  

The position of the peaks on the potential axis (Ep) is related to the formal potential of the redox process. The fonnal potential for a reversible couple is 

The separation between the peak potentials (for a reversible couple) is given by AEp = Ep, -Ep, = V (2-3) 

the peak separation ean be used to determine the number of electrons tiansferred, and as a criterion for a Nemstian behavior. Accordingly, a fast one-electron process exhibits a AEp of about 59 mV Both the cathodic and anodic peak potentials are independent of tire scan rate. It is possible to relate the half-peak potential (Epf2, where the current is half of the peak current) to the polarographic half-wave potential, Ei 2- 

---

The other parameters used in the calculation of STOMP and GIIP have been discussed in Section 5.4 (Data Interpretation). The formation volume factors (B and Bg) were introduced in Section 5.2 (Reservoir Fluids). We can therefore proceed to the quick and easy deterministic method most frequently used to obtain a volumetric estimate. It can be done on paper or by using available software. The latter is only reliable if the software is constrained by the geological reservoir model. 

Modem a bidirectional modem connection (asynchronous, 28000 b/s) between the PC of the on-site AEBIL system and a laboratory PC has enabled transfer of data files and monitor presentations from the peripheral station to the central one, strongly enhancing initial assistance in data interpretation at drastically reduced costs. 

Increasing mechanisation and automatisation of non-destructive inspection procedures lead to an increasing amount of data that is acquired in a short amount of time. This in turn creates a need for reliable automatic or automated data interpretation techniques. 

The use of automatic interpretation techniques serves first of all the increase in the speed of the data interpretation however, very important goals are also the increase in the reliability and reproducibility of the inspection. Also, in some situations, computer systems can help to interpret complex data because they are capable of analysing more data simultaneously than it can be done by an operator. 

First, the typical characteristics of inspection problems which result in heterogeneous data are presented. Next, typical AI techniques which can be used for the automated data interpretation are presented. The applicabihty of the techniques to various inspection problems is discussed. Two example apphcations for automatic NDT data interpretation are briefly described, and finally, the conclusions are given. 

Automated data interpretation will usually be done using some statistical or AI technique. Because statistical classifiers are similar in their use to neural networks [Sarle, 1994] we will not discuss them separately. 

Hybrid systems. Depending on the problem to be solved, use can also be made of a combination of techniques leading to a hybrid system. For example, a rule-based system may use neural networks for solving classification subproblems (as is described in [Hopgood, 1993]), or a combination of a rule-based and a CBR system can be used as in the system for URS data interpretation described later in this paper. 

Table 1 Influence of various factors on the use of AI techniques for NDT data interpretation.

Case-based reasoning. The main advantage of CBR systems for NDT data interpretation is that they can cope with data coming from inspection of varying constructions under varying conditions with various system settings due to their ability to learn from the data classified by the operator. In such situations no reliahle statistical classifier can be designed, and the rule-hased classifiers would be either very inefficient or unpractically complex. 

CBR systems require that the operator is present during data interpretation. The operator has to provide the correct classification for the data that could not be interpreted by the system. This means that completely automatic CBR system is not feasible, but in case of the NDT inspection operator is almost always present. 

Within the framework of the ANDES (Automatic Non-Destructive Evaluation System) at TNO Institute of Applied Physics two prototype applications have been developed for automated NDT data interpretation - both using the CBR methodology. 

Increased trust in pattern recognition The active user involvement in the data mining process can lead to a deeper understanding of the data and increases the trust in the resulting patterns. In contrast, "black box" systems often lead to a higher uncertainty, because the user usually does not know, in detail, what happened during the data analysis process. This may lead to a more difficult data interpretation and/or model prediction. 

The simplest interpretation of stm images is in terms of surface topography. However, care must be exercised in this interpretation, since in teahty, tunneling probabiUty is really measured. The many subdeties of stm data interpretation ate beyond the scope of this article. The interested reader is referred to references 14 and 15 for a more detailed discussion of these issues. 

Evidence of the appHcation of computers and expert systems to instmmental data interpretation is found in the new discipline of chemometrics (qv) where the relationship between data and information sought is explored as a problem of mathematics and statistics (7—10). One of the most useful insights provided by chemometrics is the realization that a cluster of measurements of quantities only remotely related to the actual information sought can be used in combination to determine the information desired by inference. Thus, for example, a combination of viscosity, boiling point, and specific gravity data can be used to a characterize the chemical composition of a mixture of solvents (11). The complexity of such a procedure is accommodated by performing a multivariate data analysis. 

S. N. Denting and S. L. Morgan, Experimental Design A Chemometric Approach Elsevier Science Publishing Co., Inc., Amsterdam, The Netherlands, 1987. D. D. Wolff and M. L. Parsons, Pattern Recognition Approach to Data Interpretation Plenum Press, New York, 1983. 

In the chemical engineering domain, neural nets have been appHed to a variety of problems. Examples include diagnosis (66,67), process modeling (68,69), process control (70,71), and data interpretation (72,73). Industrial appHcation areas include distillation column operation (74), fluidized-bed combustion (75), petroleum refining (76), and composites manufacture (77). 

Numeric-to-symbohc transformations are used in pattern-recognition problems where the network is used to classify input data vectors into specific labeled classes. Pattern recognition problems include data interpretation, feature identification, and diagnosis. 

Measurement Selection The identification of which measurements to make is an often overlooked aspect of plant-performance analysis. The end use of the data interpretation must be understood (i.e., the purpose for which the data, the parameters, or the resultant model will be used). For example, building a mathematical model of the process to explore other regions of operation is an end use. Another is to use the data to troubleshoot an operating problem. The level of data accuracy, the amount of data, and the sophistication of the interpretation depends upon the accuracy with which the result of the analysis needs to oe known. Daily measurements to a great extent and special plant measurements to a lesser extent are rarelv planned with the end use in mind. The result is typically too little data of too low accuracy or an inordinate amount with the resultant misuse in resources. 

Elastic recoil spectrometry (ERS) is used for the specific detection of hydrogen ( H, H) in surface layers of thickness up to approximately 1 pm, and the determination of the concentration profile for each species as a function of depth below the sample s surfece. When carefully used, the technique is nondestructive, absolute, fast, and independent of the host matrix and its chemical bonding structure. Although it requires an accelerator source of MeV helium ions, the instrumentation is simple and the data interpretation is straightforward. 

Though a powerfiil technique, Neutron Reflectivity has a number of drawbacks. Two are experimental the necessity to go to a neutron source and, because of the extreme grazing angles, a requirement that the sample be optically flat over at least a 5-cm diameter. Two drawbacks are concerned with data interpretation the reflec-tivity-versus-angle data does not directly give a a depth profile this must be obtained by calculation for an assumed model where layer thickness and interface width are parameters (cf., XRF and VASE determination of film thicknesses. Chapters 6 and 7). The second problem is that roughness at an interface produces the same effect on specular reflection as true interdiffiision. 

FIGURE 5.56 The threshold region for chronic dose-response curves. [Reprinted with permission from Tardiff, R.G., and Rodricks, J.V. (1987). (Eds.), Toxic Substances and Human Risks Principles of Data Interpretation. New York Plenum Press.]... 

W..A. Heitbrink,. M.G, Gressel, T.C. Cooper. Video exposure monitoring—,A mean.s oi saidying. sources of occupational air contaminant exposure. Part 2 Data interpretation. Ai>p . Occup. Environ. Hyg. 8(4), 199,). 

Thus, the Tsai-Wu tensor failure criterion is obviously of more general character than the Tsai-Hill or Hoffman failure criteria. Specific advantages of the Tsai-Wu failure criterion include (1) invariance under rotation or redefinition of coordinates (2) transformation via known tensor-transformation laws (so data interpretation is eased) and (3) symmetry properties similar to those of the stiffnesses and compliances. Accordingly, the mathematical operations with this tensor failure criterion are well-known and relatively straightforward. 

The overall process of data interpretation and the development of suitable remedial strategies once a set of causes has been identified, is set out in Figure 6.4. The two-stage process of confirming the initial causal hypothesis is recommended to overcome the tendency to jump to a premature conclusion and to interpret all subsequent information on the basis of this conclusion. 

FIGURE 6.4. Data Interpretation, Remedial Strategy Generation, and Implementation. 

---