Drexel dragonThe Math ForumDonate to the Math Forum



Search All of the Math Forum:

Views expressed in these public forums are not endorsed by Drexel University or The Math Forum.


Math Forum » Discussions » Software » comp.soft-sys.matlab

Topic: neural network
Replies: 3   Last Post: Jul 26, 2014 7:22 PM

Advanced Search

Back to Topic List Back to Topic List Jump to Tree View Jump to Tree View   Messages: [ Previous | Next ]
Murugan Solaiyappan

Posts: 33
Registered: 12/4/10
Re: neural network
Posted: Jul 25, 2014 10:53 AM
  Click to see the message monospaced in plain text Plain Text   Click to reply to this topic Reply

"Murugan Solaiyappan" wrote in message <lqor9e$g2s$1@newscl01ah.mathworks.com>...
Thanks for your immediate reply Mr.Greg.
Sorry for missing target data. the modified
> I want to predict the closing index of stock market. here is my data set,


Data set 1

> Input (1 to 4 column) and Target (5th column)
>
> open high low close close
> jan 1 6033.12 6067.37 6032.62 6049.89 5962.24
> jan 2 6008.59 6021.49 5935.59 5962.24 5884.24
> jan 3 5947.6 5962.46 5875.2 5884.34 5855.45
> jan 4 5912.29 5919.01 5833.02 5855.45 5800.19
> jan 5 5845.02 5850.8 5794.54 5800.19 5713.55
> jan 6 5800.92 5822.43 5706.75 5713.55 5707.16
> jan 7 5698.31 5717.23 5630.83 5707.16 5678.86
> jan 8 5737.99 5773.14 5635.84 5678.86 5625.88
> jan 9 5700.84 5716.28 5596.4 5625.88 5641.88
>


data set 2

close close
> 6049.89 5962.24
> 5962.24 5884.24
> 5884.34 5855.45
> 5855.45 5800.19
> 5800.19 5713.55
> 5713.55 5707.16
> 5707.16 5678.86
> 5678.86 5625.88
> 5625.88 5641.88


> some research articles says that , open,high,low,close is treated as input and close is treated as target. (Data set1)

>some research articles says that, previous closing stock index is treated as input and today clsoing stock index is target (Data set2)

I create a feed forward neural network architecture and I use the dataset2. Performance measurement RMSE = 67 and R=0.99 or 1.

My first question is the data preparation is correct or not. If it correct which one gives the best result. I used matlab version 2010a.


[ I N ] = size(input) % [ 4 9 ]
[ O N ] = size(target) % [ 1 9 ]

I know the size of the input and target vector. But I have doubt with the content (Please see dataset1 and dataset2) of the data set.

% Overfitting limit for hidden nodes

Hub = -1 + ceil( (N*O-O) / ( I + O + 1)) % 1

What is Hub? I can't understand.

% N too low for validation stopping. So use regularization.
% Try training with noise added data and testing with original data
% Plot performance vs signal-to-noise ratio

how to add noise data to the original data?

I need your timing help.

thanks in advance



Point your RSS reader here for a feed of the latest messages in this topic.

[Privacy Policy] [Terms of Use]

© Drexel University 1994-2014. All Rights Reserved.
The Math Forum is a research and educational enterprise of the Drexel University School of Education.