Fitting in matlab? How to and bit of theory

Dear all, I have run an experiment (for 7 different incremental inputs) with 500 runs for every different input. Totally this is 500*7 outputs.

The outputs are number of misses. I would like to do a plot with the y axis the outputs (number of misses) and as x axis the 7 different incremental inputs (30,90,150 and so on...).

I did already a nice boxplot for that different 7 sets and I also want to do a fitting that will predict the values between the 7 different sets. Could you please help me understand a) what type of fitting is suitable for results that look like decaying exponentially? b) How I can measure how good this fitting is?

I would like to thank you in advance for your help

Best Regards Alex

Answer 1

[Abhishek Murthy]

using prony fitting you can fit exponentials to your data. Google "prony fitting" or "prony estimation" for more details.

The goodness of fit can be measured by standard metrics like L2 error.