# Tutorial for multiple variational methods using Poisson regression?

**URL:** <https://discourse.edwardlib.org/t/tutorial-for-multiple-variational-methods-using-poisson-regression/101>\
**Category:** General\
**Created:** [April 28, 2017, 2:39am UTC](https://discourse.edwardlib.org/t/tutorial-for-multiple-variational-methods-using-poisson-regression/101 "2017-04-28T02:39:09Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![John-Boik](https://avatars.discourse-cdn.com/v4/letter/j/91b2a8/32.png) [@John-Boik](https://discourse.edwardlib.org/u/John-Boik)\
**Post date:** [April 28, 2017, 2:39am UTC](https://discourse.edwardlib.org/t/tutorial-for-multiple-variational-methods-using-poisson-regression/101/1 "2017-04-28T02:39:09Z")

</div>

I’m an independent researcher working on a pet project. The general idea is to use a Gaussian Process nonlinear autoregressive model to do forecasting.

I hope to work up to the larger model by building (and understanding) simpler ones. So, I thought I would start by writing a tutorial for variations on Poisson GP regression.

Perhaps such a tutorial would be useful to other new Edward users, beside myself. If anyone is interested in assisting, let me know. I would make a Jupyter notebook, and might use both AutoGrad and Edward (TensorFlow) to highlight differences and similarities. I want to make sure I understand what is happening under the hood.

For inference, I’d like to use a Gaussian mean field for q(), a normalizing flow over Gaussian mean fields, and a likelihood-free variational approach.

I have zero experience with Edward, so I’d be grateful for assistance.

John

---

<div class="post-metadata">

**Author:** ![dustin](https://yyz1.discourse-cdn.com/flex035/user_avatar/discourse.edwardlib.org/dustin/32/134_2.png) [@dustin](https://discourse.edwardlib.org/u/dustin)\
**Post date:** [April 28, 2017, 12:22pm UTC](https://discourse.edwardlib.org/t/tutorial-for-multiple-variational-methods-using-poisson-regression/101/2 "2017-04-28T12:22:26Z")

</div>

Can you elaborate on what you mean by a “Poisson GP regression” model? There are examples online of [linear models](http://edwardlib.org/tutorials/supervised-regression) and [logistic regression](https://github.com/blei-lab/edward/blob/master/examples/bayesian_logistic_regression.py). It’s easy to write a Poisson likelihood and use a log link function instead of Bernoulli-logit.

A piece of advice: Edward models are tightly integrated into the TensorFlow graph. This means instead of autograd, you can write down the model to leverage `tf.gradients` over, say, its log joint density. “Mak[ing] sure that [you] understand what is happening” could still be done in Edward. You just code up the model in Edward and not rely on one of the built-in inference engines.

---

<div class="post-metadata">

**Author:** ![John-Boik](https://avatars.discourse-cdn.com/v4/letter/j/91b2a8/32.png) [@John-Boik](https://discourse.edwardlib.org/u/John-Boik)\
**Post date:** [April 29, 2017, 2:56pm UTC](https://discourse.edwardlib.org/t/tutorial-for-multiple-variational-methods-using-poisson-regression/101/3 "2017-04-29T14:56:13Z")

</div>

Thanks Dustin. I think I need to dig into the (impressive) Edward code a bit more. But so far I am not seeing a normalizing flow inference engine. I do see the ImplicitKLqp and the others I was interested in.

I’m curious about differences/strengths/weaknesses between Autograd and TensorFlow, and am more familiar with the syntax of the former. I will try to work out a few problems for side-by-side comparisons.

John

---

<div class="post-metadata">

**Author:** ![dustin](https://yyz1.discourse-cdn.com/flex035/user_avatar/discourse.edwardlib.org/dustin/32/134_2.png) [@dustin](https://discourse.edwardlib.org/u/dustin)\
**Post date:** [May 28, 2017, 10:33pm UTC](https://discourse.edwardlib.org/t/tutorial-for-multiple-variational-methods-using-poisson-regression/101/4 "2017-05-28T22:33:57Z")

</div>

Just to reiterate: A tutorial explaining how to build a GP regression model, and how to handle all sorts of variations you might need for forecasting, would be amazing. Let me know if I can help.

---

<div class="post-metadata">

**Author:** ![John-Boik](https://avatars.discourse-cdn.com/v4/letter/j/91b2a8/32.png) [@John-Boik](https://discourse.edwardlib.org/u/John-Boik)\
**Post date:** [May 29, 2017, 2:56pm UTC](https://discourse.edwardlib.org/t/tutorial-for-multiple-variational-methods-using-poisson-regression/101/5 "2017-05-29T14:56:20Z")

</div>

Thanks much Dustin. I’ve started work on a tutorial, but it will take me some time. Once its in shape, or if I have questions, I’ll drop you a note.  
Best,  
John