Hi there, I’ve been playing around with the Edward and finding it great with the illustrative examples. In the regression tutorial Edward – Supervised Learning (Regression) it states assume σw2,σb2\sigma_w^2,\sigma_b^2σw2,σb2 are known prior variances and σy2\sigma_y^2σy2 is a known …

Not sure if you solved this already, but you need a prior with positive support. A well-behaving one for training is the log Normal prior. Also see A toy normal model failed (klqp) and why? for some discussion of priors.

Variational parameters on prior and likelihood variances

MushroomHunting August 12, 2017, 12:36am 3

Cheers;

Yep found out that was the problem along with incorrectly specifying the prior inside the likelihood model

it’s working nicely now

1 Like

Topic		Replies	Views
Unable to inference the variance in a simple bayesian linear regression	0	917	January 7, 2018
Fixed scales in regression example	6	1449	August 29, 2017
Parameters not getting updated	4	950	May 11, 2017
Plan for Hierarchical Variational Model? What is Edward inferring when there is hyperprior in linear regression?	1	859	July 16, 2018
Bayesian regression	4	1229	June 26, 2017

Variational parameters on prior and likelihood variances

Related topics