is it possible in matlab to have an aggregated target data
for example, my target data are average daily values but I want my neural network to predict hourly values since all my input is on an hourly time basis. Can I add a processing function to the output (like histc and accumarray to obtain the average daily values from hourly values) so that the NN is trained to the daily values that I have?
Below the bare bones of code of the NN I am currently using
net = fitnet(10);
[net,tr] = train(net,inputs,targets);

1 Kommentar

Greg Heath
Greg Heath am 28 Aug. 2015
How is the averaging done?
y(i) = mean(x(i-11:i)
or
y(i) = mean(x(i-6:i+5))
?

Melden Sie sich an, um zu kommentieren.

 Akzeptierte Antwort

Greg Heath
Greg Heath am 29 Aug. 2015

0 Stimmen

If you are going to try to predict hourly values, forget about your daily average data. I don't see it helping at all.
Just use the input-delay time series function
help timedelaynet
doc timedelaynet.
Hope this helps.
Thank you for formally accepting my answer
Greg

Weitere Antworten (0)

Kategorien

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

Gefragt:

am 28 Aug. 2015

Beantwortet:

am 29 Aug. 2015

Community Treasure Hunt

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

Start Hunting!

Translated by