Predictive models performance evaluation

In this section the air quality prediction models are evaluated based on the daily averaged PMio concentrations recorded at the ambient air quality monitoring station in Macao between 2001 and 2005. The station is located at the Taipa Grande Hill, which has an altitude of 159.2 m above sea level. Therefore, the measurements are representative of the general ambient air quality of Macao. To evaluate the performance of each model, some well-known performance indices are utilized and they are the root-mean-square error ... 

Numerical models are used to predict the performance and assist in the design of final cover systems. The availability of models used to conduct water balance analyses of ET cover systems is currently limited, and the results can be inconsistent. For example, models such as Hydrologic Evaluation of Landfill Performance (HELP) and Unsaturated Soil Water and Heat Flow (UNSAT-H) do not address all of the factors related to ET cover system performance. These models, for instance, do not consider percolation through preferential pathways may underestimate or overestimate percolation and have different levels of detail regarding weather, soil, and vegetation. In addition, HELP does not account for physical processes, such as matric potential, that generally govern unsaturated flow in ET covers.39 42 47... 

In summary, our data provide evidence for the suitability of zebrafish eleuther-oembryos as a predictive vertebrate model for evaluating the effect of individual chemicals and mixtures on thyroid gland function. TIQDT performed on zebrafish eleutheroembryos is an alternative whole-organism screening assay that provides relevant information for environmental and human risk assessments. 

In another work, Parra and coworkers proposed a method based on chemically modified voltammetric electrodes for the identification of adulterations made in wine samples, by addition of a number of forbidden adulterants frequently used in the wine industry to improve the organoleptic characteristics of wines, like, for example, tartaric acid, tannic acid, sucrose, and acetaldehyde (Parra et ah, 2006b). The patterns identified via PCA allowed an efficient detection of the wine samples that had been artificially modified. In the same study, PLS regression was applied for a quantitative prediction of the substances added. Model performances were evaluated by means of a cross-validation procedure. 

A way to start evaluating how a model performs when different factors are considered is to evaluate how much of the information (i.e. variance) in the X- and Y-blocks is explained. We expect that whichever the particular number of factors is the optimum, no more relevant information would enter the model after such a value. Almost any software will calculate the amount of variance explained by the model. A typical output will appear as in Table 4.1. There, it is clear that not all information in X is useful to predict the concentration of Sb in the standards, probably because of the interfering phenomena caused by the concomitants. It is worth noting that only around 68% of the information in X is related to around 98% of the information in Y ([Sb]). This type of table is not always so clear and a fairly important number of factors may be required to model a large percentage of information in X and, more importantly, in Y. As a first gross approach, one can say that the optimal dimensionality should be... 

Probably the most common internal validation method, cross-validation, involves the execution of one or more internal validation procedures (hereby called sub-validations), where each procedure involves the removal of a part of the calibration data, use of the remaining calibration data to build a subset calibration model, and subsequent application of the removed data to the subset calibration model. Unlike the Model fit evaluation method discussed earlier, the same data are not used for model building and model testing for each of the sub-validations. As a result, they can provide more realistic estimates of a model s prediction performance, as well as better assessments of the optimal complexity of a model. 

The following description of several algorithms and software is partly based on the comparative evaluation of model performance by Rorije et al. (1999) and the recent review of broadly applicable methods for predicting biodegradation by Jaworska et al. (2003). Table 14.2 summarizes the main features of some models. 

During the approval process, the regulatory agencies can use models to check that the tests are adequately performed. The predictive modeling approach is strongly recommended by the American Food and Drug Administration that already uses mathematical models in its evaluation of applications for drug approval. 

If one has intrinsic and apparent reaction kinetics available, then Equation 10 may be viewed as a three-parameter model (t)ce, b wo> B -d f°r prediction of isothermal trickle-bed reactor performance. However, Biwo depends on two mass transfer coefficients and a priori model parameter evaluation is no simpler than before. 

Classification model performance is evaluated by classification parameters, both for fitting and predictive purposes. 

---