SDNEURAL Feed-forward neural network [P,ETR,ETS]=SDNEURAL(DATA,options) INPUT DATA training dataset OPTIONS units number of units in hidden layer (def: 10) epochs number of epochs (def: 1000) rate training rate (def: 0.3) momentum momentum term (def: 0.8) valfrac fraction of DATA used for validation (def: 0.25) targets target matrix with the same number of samples as DATA OUTPUT P neural network pipeline ETR,ETS mean square error on training set and on test set DESCRIPTION SDNEURAL implements training of feed-forward neural network with one hidden layer. Mean square error (MSE) is minimized with respect to the validation set (by default 25% of DATA). SDNEURAL may approximate the user-supplied target matrix using the 'targets' option. READ MORE http://doc.prsdstudio.com/latest/guide/classifiers.html#sdneural REFERENCES * "Elements of Artificial Neural Networks", K.Mehrotra, C.K.Mohan, S.Ranka * Genevieve Orr: http://www.willamette.edu/~gorr/classes/cs449/intro.html * D.Patterson: http://www.csee.umbc.edu/~dpatte3
sdneural is referenced in examples: