Wrapper for neural network ensembles implemented in alglib

[gkronber]

r6577: added first implementation of neural networks for regression wrapper for alglib.

[gkronber]

r6578: added parameters for neural network regression algorithm.

[gkronber]

r6579: implemented neural networks for classification.

[gkronber]

r6580: added implementations for regression and classification with neural network ensembles (wrappers for alglib).

[mkommend]

Reviewed the source of neural network data analysis algorithm. During testing it noted that the default parameter values are quite different for neural networks and neural ensemble networks, and I wonder if this was on purpose. Additionally the ensemble network parameter values yield to rather poor results (tested on poly-10 and default classification problem), compared to standard neural networks.

[gkronber]

r6718: fixed bug in alglib that resulting in random models when using the early stopping training method for multi layer perceptrons.

[gkronber]

r6719: removed range limit for MLP output

[gkronber]

After r6718:r6719 the performance of both algorithms, normal MLP and MLP ensemble, are comparable.

[gkronber]

r6720: added (conditional) parameter hiding