neural network to detect points in signal

1 view (last 30 days)
i have two signals they look like sine wave but with some noise . in this signals of length 100 each i have 1-5 points each time(caqn change) i want to detect. my question is can i use neural network to detect points. i want network will have 2* 100 inputs which represent signal and 100 outputs which can be 1 or 0. the issue that each time i have different signals and the point location is different is it possible to design such network

Accepted Answer

Greg Heath
Greg Heath on 28 Mar 2016
You don't need the complexity of a neural network.
Just use something like
d = 5
y = zeros(1,100);
for i = 1:100-d
if sae(x1(i:i+d)-x2(i:i+d)) > thresh
y(i+d) = 1
end
end
Hope this helps.
Thank you for formally accepting my answer
Greg

More Answers (1)

michael scheinfeild
michael scheinfeild on 29 Mar 2016
more complex example . the marked lines are outputs as one. they represent some stage of system which marked manually as ground truth. if i use if then else condition's it will be to large
  1 Comment
Greg Heath
Greg Heath on 30 Mar 2016
I don't fully understand your comment. However if you are complaining that multiple outputs have to be compared to multiple ground truth standards, then I feel for you.
Good Luck,
Greg

Sign in to comment.

Community Treasure Hunt

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

Start Hunting!