Virtual Computational Chemistry Laboratory

Input data Output results Example List of key words


Keyword of Integer Type

This keyword indicate number of network in ensemble that will be analyzed. This allows to estimate variance of neural network predictions for the whole data set  and for each data entry.

The result of  calculation is simple average of all networks. The program also estimates variance of RMSE of all networks in the ensemble. It recalculates neural networks with RMSE error that is higher than 3 standard deviations from the average RMSE.

We recommend to use ENSEMBLE of 100-400 neural networks. A smaller number provides a high error of neural network results, while a larger number requires too long calculations.

The range of this keyword is [1,10000] and default value is {200}.

See FAQ if you have questions. How to cite this applet? Are you looking for a new job in chemoinformatics?

Copyright 2001 -- 2016 All rights reserved.