Machine learning parameters

Several nonlinear QSAR methods have been proposed in recent years. Most of these methods are based on either ANN or machine learning techniques. Both back-propagation (BP-ANN) and counterpropagation (CP-ANN) neural networks [33] were used in these studies. Because optimization of many parameters is involved in these techniques, the speed of the analysis is relatively slow. More recently, Hirst reported a simple and fast nonlinear QSAR method in which the activity surface was generated from the activities of training set compounds based on some predefined mathematical functions [34]. 

Neural Nets (NNs) relate a set of input neurons with an output neuron (providing the prediction label of a data point) by a network of layers of neurons in the interior. They are certainly among the most frequently used Machine Learning methods in the field [148] and allow for a high degree of customization since the architecture of the network itself is part of the parameters the user may define. 

Predictive models are built with ANN s in much the same way as they are with MLR and PLS methods descriptors and experimental data are used to fit (or train in machine-learning nomenclature) the parameters of the functions until the performance error is minimized. Neural networks differ from the previous two methods in that (1) the sigmoidal shapes of the neurons output equations better allow them to model non-linear systems and (2) they are subsymbolic , which is to say that the information in the descriptors is effectively scrambled once the internal weights and thresholds of the neurons are trained, making it difficult to examine the final equations to interpret the influences of the descriptors on the property of interest. 

Expert feedback mainly consists of correct mappings between the schemas to be matched. These mappings can be seen as a bootstrap for the schema matcher, i.e., knowledge is taken as input by machine learning algorithms to classify schema instances. It may be a compulsory parameter such as in LSD/Glue [Doan et al. 2001, 2003] and APFEL [Ehrig et al. 2005],... 

The second category of matchers use machine learning techniques to combine similarity measures. However, they share almost the same parameters than the first category. 

In this section, we have mainly presented user inputs, i.e., optional preferences and parameters applied to data. To sum up, the quality can be improved by using external resources and expert feedback. Several tools are based on machine learning techniques either as a similarity measure (mostly at the instance level) or as a means of combining the results of similarity measures. In both cases, training data is a crucial issue. Finally, many tools propose preferences or options which add more flexibility or may improve the matching quality. The next section focuses on the parameters at the similarity measure level. 

The experiments are taken with Weka machine learning package, and the results are evaluated by several statistical metrics. We provide an explanation of these parameters in the following parts. The statistical metrics of the predictive results are evaluated by accuracy, Kappa statistics, mean absolute error, root mean squared error, and relative absolute error. For the prediction performance, we normally refer to accuracy, which is the... 

---