Extract the equation of the neural network

32 views (last 30 days)
Hi guys, please help me with neural network topic, can you extract the equation of neural networks Matlab, to be use in Excel (read the whole equation) and publish in a slideshow, yes or no and why (documentation).
There is a document which can be read to see if it is possible or not. For what serving this equation "(y = b2 + LW * tansig (b1 + IW * x)", i read for the weight but can you teach me.
And, there is a neural network program that can deliver the equation. My job consists 9 variables of input and 1 output, trying to relate events of nature, precipitation -》soil erosion.
pleaseeeeee help me, i read a lot topic about neural network but anyone can say explicit and educative way the solution.
Pleaseeeeeeeee i need help

Accepted Answer

Greg Heath
Greg Heath on 8 Oct 2015
There are 7 basic NNs in the NN Toolbox. Although they all have default configurations, very often modifications are used. I don't know of any program that will automatically yield the analytic input/output relationship of an arbitrary net.
The command
view(net)
will yield a visual representation, from which a non-beginner can piece together the analytic representation for a nonstandard configuration.
However, there doesn't seem to be an easy out for the beginner.
My recommendation is to first learn the configurations and corresponding analytic representations for the following
1. FITNET
2. PATTERNNET
3. TIMEDELAYNET
4. NARNET
5. NARXNET
Then you will be better prepared to deal with the less standard configurations.
Hope this helps.
  1 Comment
Greg Heath
Greg Heath on 8 Oct 2015
The final 2 of the 7 I mentioned are
FEEDFORWARDNET - Basic multilayer perceptron (MLP) that is
called by FITNET and PATTERNNET
SELFORGMAP - Self organizing map for clustering data into
similarity groups. Useful for visualizing multidimensional
data in a 2 dimensional representation.
Hope this helps.
Greg

Sign in to comment.

More Answers (1)

Walter Roberson
Walter Roberson on 7 Oct 2015
It appears to me that the equations that can be derived are difficult to interpret; see http://www.mathworks.com/matlabcentral/newsreader/view_thread/62107

Categories

Find more on Deep Learning Toolbox in Help Center and File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!