Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
NCTM or The Math Forum.



Re: neural network
Posted:
Jul 24, 2014 1:25 PM


"Murugan Solaiyappan" wrote in message <lqor9e$g2s$1@newscl01ah.mathworks.com>... > Dear matlab friends, > > Sorry friends. previous post is sent by incompletely. > > I want to predict the closing index of stock market. here is my data set, > > Input (1 to 4 column) and Target (5th column) > > open high low close > 6033.12 6067.37 6032.62 6049.89 > 6008.59 6021.49 5935.59 5962.24 > 5947.6 5962.46 5875.2 5884.34 > 5912.29 5919.01 5833.02 5855.45 > 5845.02 5850.8 5794.54 5800.19 > 5800.92 5822.43 5706.75 5713.55 > 5698.31 5717.23 5630.83 5707.16 > 5737.99 5773.14 5635.84 5678.86 > 5700.84 5716.28 5596.4 5625.88 > > research articles says that , open,high,low,close is treated as input and close is treated as target. > > My question is input contains 9 element and the target contains either from 1 to 9 or 2 to 9. or anything else. > > Please briefly explain to me regarding input and target pair for closing stock index prediction. I expect your answer as early as possible?
You forgot to list the target column
[ I N ] = size(input) % [ 4 9 ] [ O N ] = size(target) % [ 1 9 ]
% Overfitting limit for hidden nodes
Hub = 1 + ceil( (N*OO) / ( I + O + 1)) % 1
% N too low for validation stopping. So use regularization. % Try training with noise added data and testing with original data % Plot performance vs signaltonoise ratio
help trainbr doc trainbr
Hope this helps.
Greg



