Projects authored by daniel steinberg.


Logo Aboleth 0.7

by dsteinberg - December 14, 2017, 02:39:19 CET [ Project Homepage BibTeX Download ] 2631 views, 777 downloads, 3 subscriptions

About: A bare-bones TensorFlow framework for Bayesian deep learning and Gaussian process approximation

Changes:

Release 0.7.0

  • Update to TensorFlow r1.4.

  • Tutorials in the documentation on:

  • Interfacing with Keras

  • Saving/loading models

  • How to build a variety of regressors with Aboleth

  • New prediction module with some convenience functions, including freezing the weight samples during prediction.

  • Bayesian convolutional layers with accompanying demo.

  • Allow the number of samples drawn from a model to be varied by using placeholders.

  • Generalise the feature embedding layers to work on matrix inputs (instead of just column vectors).

  • Numerous numerical and usability fixes.


Logo revrand 1.0.0

by dsteinberg - January 29, 2017, 04:33:54 CET [ Project Homepage BibTeX Download ] 18134 views, 3818 downloads, 3 subscriptions

Rating Empty StarEmpty StarEmpty StarEmpty StarEmpty Star
(based on 1 vote)

About: A library of scalable Bayesian generalised linear models with fancy features

Changes:
  • 1.0 release!
  • Now there is a random search phase before optimization of all hyperparameters in the regression algorithms. This improves the performance of revrand since local optima are more easily avoided with this improved initialisation
  • Regression regularizers (weight variances) associated with each basis object, this approximates GP kernel addition more closely
  • Random state can be set for all random objects
  • Numerous small improvements to make revrand production ready
  • Final report
  • Documentation improvements

Logo libcluster 2.3

by dsteinberg - February 27, 2016, 00:36:01 CET [ Project Homepage BibTeX BibTeX for corresponding Paper Download ] 8033 views, 1587 downloads, 3 subscriptions

About: An extensible C++ library of Hierarchical Bayesian clustering algorithms, such as Bayesian Gaussian mixture models, variational Dirichlet processes, Gaussian latent Dirichlet allocation and more.

Changes:

New maximum cluster argument for all algorithms. Also no more matlab interface since it seemed no one was using it, and I cannot support it any longer.


Logo linearizedGP 1.0

by dsteinberg - November 28, 2014, 07:02:54 CET [ Project Homepage BibTeX BibTeX for corresponding Paper Download ] 3681 views, 758 downloads, 1 subscription

About: Gaussian processes with general nonlinear likelihoods using the unscented transform or Taylor series linearisation.

Changes:

Initial Announcement on mloss.org.