ABC-PRC: Correction posted on PNAS

In connection with our ABC-PMC paper, to appear in Biometrika, the authors of the original ABC-PRC paper have recently published a corrigendum in PNAS that agrees with our analysis of the bias., due to the likelihood ratio approximation based on two unbiased estimators being itself biased. The correction they propose is to modify the denominator in an importance sampling manner, as in ABC-PMC. So this kind of nicely solves the earlier controversy surrounding those papers!

This entry was posted on September 4, 2009 at 4:16 pm and is filed under Statistics with tags ABC, Bayesian statistics, simulation. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.

2 Responses to “ABC-PRC: Correction posted on PNAS”

Inference in epidemic models w/o likelihoods « Xi'an's Og Says:
July 21, 2010 at 12:11 am

[…] would further the improvement brought by sequential methods. (The authors were aware of the correction of Sisson et al., and used instead the version of Toni et al. They also mention the arXived notes […]

Reply
Julien Says:
September 9, 2009 at 8:38 pm

Long life to “A whole happy ABC family again”, as said Mark Briers at ABC in Paris !

Reply

Xi'an's Og