I am trying to estimate effective population size (Ne) and changing time with microsatellite data in DIYabc 2.1.0. After I ran the analysis after generating simulated data set, I find the parameter estimation sometimes looks strange. For example, I could find Ne1 is smaller than Ne2 in a specific scenario with the greatest possibility even if I define Ne1 < Ne2 in prior setting of this scenario.
Another common problem is that the posterior curve is not looking good. Estimated Ne can be very small or large and out of the range in prior setting.
I am wondering why this happens and how to adjust my prior setting to fix it?
Can someone can help answer this question? Thank you for your time and help!
You can use the machine learning to estimate the parameters. Many methods are available...try to use that will definitely solve your problem. Some basic algorithm is:
The Levenberg–Marquardt algorithm is a technique that has been used for parameter extraction.
- Badges
- Science topic