GSoC‘20 Final Report

Major Pull Requests

  • pymc-devs/pymc4#235: Implemented the basic API structure of Gaussian Processes. Implemented Latent Gaussian Process Model. Created a notebook explaining them.
  • pymc-devs/pymc4#272: Implemented/Refactored the covariance functions API for Gaussian Processes. Introduced multiple new features on top of Tensorflow Probability’s PSD API.
  • pymc-devs/pymc4 …
more ...

Marginal GP model in PyMC4

Marginal Gaussian Process

The Model

A Marginal Gaussian process jointly represents the data as a large probability distribution. This distribution, as we will see, turns out to a normal distribution in classical Bayesian approaches. Suppose we have some data \(X, y\) using which we want to predict the distribution over …

more ...

GSoC‘20 Phase 2 Report

Tasks Completed

Week 5: #303: Fix docs in GP submodule

Small fixes in the documentation suite of GP submodule so that pytest passes

Week 6: #304: fix docs of kernel functions

Fixed typos in the documentation suite of kernel functions. One inconsistency still remains which I have planned to remove …

more ...

GSoC‘20 Phase 1 Report

Tasks Completed

Week 1: #235: [MRG] ENH: Add Basic Gaussian Process Interface

Blog: GSoC Week 1 - Latent GP model and Covariance functions!

  • [x] Create a base class for kernel/covariance functions.
  • [x] Create an API for combining covariance functions.
  • [x] Create a base class for mean functions.
  • [x] Create an …
more ...

Kernels for GP Modelling in PyMC4.

Kernels for GP Modelling in PyMC4.

In this tutorial, we will explore the covariance functions aka kernels present in pm.gp module and study their properties. This will help choose the appropriate kernel to model your data properly. We will also see the semantics of additive and multiplicative kernels.

import …
more ...