General


How to Set Up the Forum as a Mailing List (3)

You might prefer to interact with the Forum as a mailing list. To follow the mailing list: Sign up on the Forum. Go to your user preferences and activate Mailing list mode. To create a new thread: Write to edwar…

Welcome to the Edward Forum (3)

Welcome to the Edward Forum. It is a haven for all things Edwardian—ranging from help with installation, coding and debugging in Edward, and any discussion about probabilistic modeling. For more community resources: …

Acceptance Rate 0 for HMC in IRT models (1)
Re-using a saved Bayesian neural network (1)
Edward: Memory leak with iterations (2)
Forming conditional distributions for inference (1)
Edward2 vs pyro (2)
Error encountering while implementing tensorflow (1)
Error of distributed training with Edward (1)
LDA with collapsed Gibbs Sampling (1)
Problems when building a variational model with MIXED GAUSSIAN prior rather than NORMAL prior (1)
Black box alpha divergence minimization (8)
Creating and running an edward model in a loop without memory leak (2)
Only inferencing partial latent variables (3)
There is error in Edward exmple (1)
ValueError: 'transform' does not handle supports of type 'categorical',how to fix the problem? (1)
Key-value pair in data does not have same shape: (2,), (500, 2) (1)
The result isn't right,someone give me advices (1)
In the Edward example unsupervised,changing ed.Gibbs to ed.SGLD,it gives error! (1)
Edward for Sequential Importance Resampling Particle Filter (8)
Why do call the codes gibbs? (1)
Someone use ed.SGLD (1)
Change ed.KLqp to ed.sgld,then it gives error.why? (1)
Edward example code Mixture (1)
Building a model to predict a proportion/probability given multiple rows/groups and potentially multiple trials per candidate (1)
Getting started (5)
Status of Edward given Edward2 / TensorFlow Probability (3)
First Edward program (2)
Suitable Distribution for variance as a Hyperprior (5)
q()=PointMass() in ed.KLqp (3)