These metrics are part of a set of very traditional statistic metrics used for evaluating CAD systems.
Then, designed PCA models are applied to new data to calculate statistic metrics and different discriminant analysis is conducted to determine whether and which fault occurs.
The results, based on different statistic metrics considering 5-fold cross validation (93.1% of precision, 95.8% of sensitivity and 91.5% of accuracy), demonstrated the effectiveness of the proposed technique based on ANNs to predict corrosion behaviour by automatic way, not requiring the use of electrochemical techniques.
In addition, natural image statistic (NSS) metrics are promising [4].
Apart from the parametric statistic, nonparametric correlation metrics such as the χ test, Spearman's ρ, and Kendall's τ are proposed, and those metrics can be applied to problems of diverse nonnormal distributions [ 23].
The types of statistical indicators, metrics, metadata, and graphics are stored, respectively, in the dimension tables Statistic Type, Metric Type, Metadata Type, and Graphic Type.
The asymptotic connection between the standard log-likelihood ratio statistic and the χ metrics used for multi-locus analysis of LD by Hill (1975) and more recently by Zhao et al. (2005) is given, for example, in Agresti (2002).
Figure 18 illustrates the KS test statistic values of both quality metrics (NFIQ and the presented one).
In almost all of these simulations, the informativeness metric accurately estimated the correct number of clusters simulated, with the only other metric, the gap statistic demonstrating similar consistent performance, while the remaining metrics performed on average, quite poorly.
As a result, these metrics (AUROC, C-statistic, specificity, likelihood ratio) provide incomplete information and can lead to overestimating the benefits of an EWS or underestimating the cost in terms of clinical resources.
However, we thought that for particular measures that were used, e.g., as test statistic for hypothesis tests or distance metrics for clustering, it might be interesting to quantify these differences more specifically.
