Sharing posts on TF Probability

[Nipun Batra]

Hi everyone

I was learning TF Probability and made a few blog posts. I wanted to share them with the community, seek feedback to improve.

 I'd be happy to also contribute some learnings from these to the TFP docs. As an example for the PCA post, I think it might be useful to add:

1) MLE

2) Add some discussion text on conditional sampling used

3) Add the graphical models

4) Add a uniform prior and show the same result as MLE

My posts (some incomplete, some a little terse) are on my blog:

1. Coin Toss (MLE, MAP, Full Bayesian)

2. Implementing variational linear regression from scratch (with reparameterization, ELBO, a lot of details on KL-divergence) 

3. Linear regression using JointDistributionCoroutineAutoBatched

4. A simple PGM in TFP - sprinkler, rain

5. Sampling from univariate and multivariate normal using standard normal (useful in reparameterization)

--

Nipun Batra,

Assistant Professor, CSE,

IIT Gandhinagar,

https://nipunbatra.github.io/