Probability: Feature Request: Maximum A Posteriori (MAP) estimation

Created on 16 May 2018  路  1Comment  路  Source: tensorflow/probability

For mixed models, it would be great if MCMC could be initialized with a more accurate estimation. This was a feature in Pymc and Edwardlib, but does not appear to be available in either tensorflow_probability or the edward2.

picture matthew-mcateer

Most helpful comment

Is there any interest in this in tfp? I鈥檓 currently working on the Stochastick Block Model example and I ended up re-implementing edward鈥檚 MAP. That would be a nice by-product.

picture rlouf on 11 Jun 2018
👍6

>All comments

Is there any interest in this in tfp? I鈥檓 currently working on the Stochastick Block Model example and I ended up re-implementing edward鈥檚 MAP. That would be a nice by-product.

rlouf picture rlouf on 11 Jun 2018
👍6
Was this page helpful?
0 / 5 - 0 ratings

Related issues

fdtomasi picture fdtomasi  路  4Comments
AliKhoda picture AliKhoda  路  5Comments
dionman picture dionman  路  5Comments
jinPrelude picture jinPrelude  路  5Comments
philbinj picture philbinj  路  4Comments