Importance ranking

The Important Sequence Model module does sensitivity studies and importance rankings for about a thousand highest frequency sequences. The analyst zooms to the most frequent plant damage category, to the most frequent sequences in that category, to the most important top event, to the most important split fraction, and to the most important cutsets. If sensitivity analysis is needed on the model as a whole, a menu option, "CLONE a Model," makes a copy of the model, c hange,s are made, and results compared. 

Table 13.5 summarizes the priority setting results for the two data sets using the NCTR Four-Phase system. When only the Phase I and II protocols are used, the system dramatically reduced the number of potential estrogens by some 80 to 85%, demonstrating its effectiveness in eliminating these most unlikely ER binders from further expensive experimentation. The Phase III CoMFA model further reduces the data size by about 5 to 10%. More importantly, the quantitative binding affinity prediction from Phase III provides an important rank-order value for priority setting. 

Here s what you should do Reread the two paragraphs about mandatory school uniforms. Decide which author you agree with most. Then, look carefully at the effects the author predicts. Which effect do you think is most important Which is least important Rank these effects in order of importance. Then, decide whether you want to start with the most important idea and end with the least important, or vice versa, start with the least important idea and end with the most important. Finally, put it all together in a paragraph in the space provided. 

TABLE 1 Importance Rankings for the Top Links from Figure 2... 

There are two different types of relationship diagrams a bubble type and a layout type. The bubble type is done first. This relationship diagram, as shown in Figure 4, uses labeled circles connected by lines to help visualize the important links. The number of lines connecting the different circles corresponds to the importance rank of the links as determined in Table 1. The primary purpose of this type of relationship chart is to help determine which work center(s) should be centrally located, which can be done simply by counting the number of lines connected to the circle. The work center(s) with the greatest number of lines should be centrally located. For simpler drawings, this step can be eliminated, in which case the layout-type relationship chart is drawn directly from Table 1. 

The Pareto Principle expresses itself in an industrial context in the form of importance ranking. In the second example (Chapter 1), the electrically driven pump, P-IOIB, has a probability of failure to start of 0.1, i.e., it will not start one times in ten. Management may decide that this failure rate is too high they wish to reduce the rate to lower than 0.02, i.e., one time in 50. 

The number of occurrences of each of the six steps listed above is shown in Table 15.1 and Figure 15.1. The events are sorted by importance ranking. 

Uniform changes Consider now the scenario HI, the results of DlMs in the directions Qx,(i = 1,2, 3) are obtained by using the analytical expression (Equation 10). The results in Table 3 show that the contribution of Cl on the total variation of system availability is more important than that of C2 or C3, and the components importance ranking is Cl > C2 > C3. 

According to these importance measures, the components importance ranking could be drawn. More precisely, C1 becomes the less important one, and the most important component is C2. 

Note however that the components ranking that we obtain are not absolute importance rankings, but the ranking relative to the DIM criterion with a specific scenario of changes. Obviously, different rankings could be obtained if a different importance measure was used. 

The results of DlM(Q,)(i = 1,3,4) for the case HI (uniform changes) are reported in Table 5. According to DlM(Q,) s measures, state 1 is the most important and state 3 is more important than state 4. This importance ranking still holds for the case where the components failure rates change proportionally (case H2), see Table 6. 

Under the ranking method devised by Jean Charles de Borda, (an eighteenth century French physicist) the DM is asked to rank the p criteria from the most important (ranked first) to the least important (ranked last). The criterion... 

Daemi, T., Ebrahimi, A. Fotuhi-Firuzabad, M. (2012). Constructing the Bayesian Network for components reliability importance ranking in composite power systems. International Journal of Electrical Power Energy Systems, 43(1) 474-480. 

Butler, D A. 1979. A complete importance ranking for components of binary coherent systems with extensions to multi-state systems. Naval Research Logistics 4 565-578. 

The exclusion of xmimportant component functions has several advantages. Firstly, since the error of the Monte Carlo integration controls the accuracy of the RS-HDMR expansion, it is possible that the inclusion of uimecessary terms can increase the integration error, reducing the accuracy of the HDMR metamodel. Secondly, if the metamodel were to be used for subsequent analysis, the lower number of terms aids its computational efficiency. The exclusion of component functions also provides an immediate level of complexity reduction, before parameter importance ranking has been performed. 

Tarantola (2008). Different importance rankings were found depending on whether moment-independent or variance-based techniques were used. The enthalpy of reaction was found to influence the entire output distribution the most, whereas its variance was most influenced by the Semenov number. However, non-important model inputs were found to be the same in both cases. Hence, if the aim of the study were to identify important and non-important parameters for the purpose of model improvement, the outcome would have been the same. 

---