# how to calculate mean square error?

11 views (last 30 days)
studentmatlaber on 13 Jun 2021
Edited: Sulaymon Eshkabilov on 11 Jul 2021
hello everyone, I separated the signals from noise by passing them through a low pass filter. I extracted the histogram of the noiseless signals. I tried to fit in matlab. the optimal distributions seem to be exponential and weilbull. I want to calculate the mean square error for these two distributions. I will use whichever distribution has the least error rate. But I don't know how to calculate mean square error. I would be very happy if you could help me with this subject.
signal=1x512712 double

Sulaymon Eshkabilov on 13 Jun 2021
Edited: Sulaymon Eshkabilov on 13 Jun 2021
Here is the sample code how to compute MSE from the histogram fit model:
x = .... % Your data. You've got it.
% Handle of the figure is needed
HH = histfit(x,Nbins); % HISTOGRAM fit model build. You've got it
%% Separate out the data from HISTOGRAM Plot and Fit Model
Data = get(HH(1),'XData'); % HIST Data
FM = get(HH(2),'YData'); % FIT Model
MD = mean(DATA); % MEAN Of Bins
FM2 = mean(reshape(FM, 2, [])); % Mean Of Bin Edges Fit
ERR = MD - FM2; % ERR
SSE = sum(ERR.^2); % Sum-Squared Error
MSE = mean(ERR.^2); % Mean-Squared-Error
studentmatlaber on 11 Jul 2021
I'm so sorry, I didn't realize I didn't hit the accept button. I used your solution for auto fit but couldn't adapt it for manual fit.