From: "Greg Heath" <>
Newsgroups: comp.soft-sys.matlab
Subject: Re: NARX learning
Date: Sat, 30 Mar 2013 05:05:05 +0000 (UTC)
Organization: The MathWorks, Inc.
Lines: 15
Message-ID: <kj5rq1$l58$>
References: <kc0c6h$l2e$> <kc0edo$s3p$>
Reply-To: "Greg Heath" <>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
X-Trace: 1364619905 21672 (30 Mar 2013 05:05:05 GMT)
NNTP-Posting-Date: Sat, 30 Mar 2013 05:05:05 +0000 (UTC)
X-Newsreader: MATLAB Central Newsreader 2929937
Xref: comp.soft-sys.matlab:792302

"Mehdi " <> wrote in message <kc0edo$s3p$>...
> "Mehdi " <> wrote in message <kc0c6h$l2e$>...
> > How to train NARX successfully?
> I trained NARX opened form successfully, Then closed it and exert the same training data to see the the Responses. The Responses are awful and bad. I dont know why?(Because I trained opened loop with enough datasets e.g 100000 and mse=10e-7)
> To solve the problem I try to train closed loop using lm. But because of large training data the training speed was very very low.I decreased data up to 1000 pairs. The training starts but after some iteration it stoped with the Maximum Mu Reached. I tested the resulted network with same data set that I have used for training. But the responses was awful again.
> Any comment will be so helpfull.

1. Are you sure the data is stationary?

2. Do not randomize the data division.

Hope this helps.