A MATLAB Central Contributor Question Sent On:Feb 02/03/16 11:26 PM
Hi Greg, I'm in a senior semester at university and our senior project is about neural network by matlab and my mission is elman back propagation neural network, but unfortunately, I don't know any thing about it. can you help me please?
1. Start with the command line help and doc documentation
lookfor elmannet
help elmannet
doc elmannet
2. I am satisfied when
a. the mean-square-error mse(target-output) is at least 100 times
smaller than the mean-target-variance mean(var(target',1))
b. the total number of weights net.numWeightElements is as small
as possible
3. Run the help and doc examples
a. Use defaults on 10 examples (Ntrials = 10, for i=1:Ntrials ...),
with different random initial weights
b. See if the number of hidden nodes can be reduced without
violating 2a
c. See if increasing the number of feedback delays makes a
significant difference
3. For references search
answers mathworks newsgroup
elman 22 26 51
elman greg 11 11 16
4. Post an attempt on one or more of the other time-series example datasets from
help nndatasets
doc nndatasets
Hope this helps.
Greg

4 Kommentare

John D'Errico
John D'Errico am 8 Feb. 2016
What is the question here?
Greg Heath
Greg Heath am 9 Feb. 2016
Bearbeitet: Greg Heath am 9 Feb. 2016
@John
I interpreted the question to be
How do I design and use an ELMAN net?
OH! ...
Did you think I should have posted this in the NEWSGROUP instead?
Greg
Explorer
Explorer am 9 Feb. 2016
If I were you, I would send via contact option in profile.
Greg Heath
Greg Heath am 11 Feb. 2016
I make it a rule to only contact requesters via the NEWSGROUP or ANSWERS. However, the subject has to be the type that I usually consider on those groups.
Greg

Melden Sie sich an, um zu kommentieren.

 Akzeptierte Antwort

Weitere Antworten (0)

Kategorien

Mehr zu Deep Learning Toolbox finden Sie in Hilfe-Center und File Exchange

Community Treasure Hunt

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

Start Hunting!

Translated by