If I have an agent-based model and I want to infer the parameters, I would normally used ABC (approximate bayesian computation), but I was recently working with someone who was using GA (genetic algorithm) for a very similar problem - which method is best and how do they compare? I think the ABC will give more information on the variance? Under which circumstances would you choose which method? The output from the collaborator is just time-series data, which goes into the GA, but I would choose more summary statistics for the ABC? Approx. 5 parameters to learn

The GA works by minimising the KL divergence of the time series

  • $\begingroup$ Can you edit the question to spell out the acronyms? I'm not sure what ABC is. I don't think there is likely to be a general answer about which is better without any information on what problem you are solving. $\endgroup$ – D.W. Feb 27 at 7:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.