False Alarm Rate Metrics

To provide a method for the evaluation, or at least comparison between two analyzers probabilities of false alarms, a model for the prediction of FAR was devised, which is based on relating the total number of the analyzer s orthogonal measurement channels and the analyzer s signal-to-noise ratio (R, ) tio to the probability of analysis errors obtained imder specific test conditions, which also corresponds to errors predicted by ROC curves [8], as illustrated by the example in Fig. 9.3.14. The area above the ROC curve, 1-Az, represents the total instrumental error of the involved analyzer, and may be plotted by simply using the result of tests yielding values of FPF (False Positive Fraction) and FNF [16]. One set of specific test conditions assumed in early model computations consisted of a simple symmetry for FPF and FNF, leading to the symmetrical Decision Matrix and Proportions shown in Fig. 9.3.14. 

Starting with a simple analyzer composed of an array of polymer film sensors, it is agreed among some researchers that, despite the inclusion of 10-20 different polymer films in the array, its maximum OCC, or n, is 5 to 6. Next, a microGC might be able to achieve a peak capacity of n = 50, if the time and energy constraints in this MGA Program were relaxed. 

Similar arguments can be made for the remaining analyzers in Table 9.3.1, taken individually (Y = 1), i.e., without the benefit of networking. Networking just 10 MG As may boost the OCC by a factor of 100 because of the intelligence derived above mere redundancy, such as associated with wind direction and speed. Such considerations generated the three sets of curves of Fig. 9.3.15. 

TCD = Thermal Conductivity D., MDID = MicroD is charge Ionization D., FD = Fluorescence D., CID = Chemi-Impedance D. (polymer film), MDOD = MicroDischarge Optical D., ITMS = Ion Trap MS, 

Note that only two parameters of the Pfp curves needed adjustment to achieve the match with the 1-Az curves obtained under the symmetrical test conditions mentioned above. 

---