A Year of Approximate Inference: Review of the NIPS 2015 Workshop

[dropcap]Probabilistic[/dropcap] inference lies no longer at the fringe. The importance of how we connect our observed data to the assumptions made by our statistical models—the task of inference—was a central part of this year's Neural Information Processing Systems (NIPS) conference: Zoubin Ghahramani's opening keynote … Continue reading A Year of Approximate Inference: Review of the NIPS 2015 Workshop

Machine Learning Trick of the Day (5): Log Derivative Trick

Machine learning involves manipulating probabilities. These probabilities are most often represented as normalised-probabilities or as log-probabilities. An ability to shrewdly alternate between these two representations is a vital step towards strengthening the probabilistic dexterity we need to solve modern machine learning problems. Today's trick, the log derivative trick, helps … Continue reading Machine Learning Trick of the Day (5): Log Derivative Trick

Machine Learning Trick of the Day (4): Reparameterisation Tricks

Our ability to rewrite statistical problems in an equivalent but different form, to reparameterise them, is one of the most general-purpose tools we have in mathematical statistics. We used reparameterisation in all the tricks we explored in this series so far: trick 1 re-expressed a … Continue reading Machine Learning Trick of the Day (4): Reparameterisation Tricks

A Statistical View of Deep Learning (II): Auto-encoders and Free Energy

With the success of discriminative modelling using deep feedforward neural networks (or using an alternative statistical lens, recursive generalised linear models) in numerous industrial applications, there is an increased drive to produce similar outcomes with unsupervised learning. In this post, I'd like to explore the connections between denoising auto-encoders as a leading approach for unsupervised learning in deep learning, and density estimation in statistics. The statistical view I'll explore casts learning in denoising auto-encoders as that of inference in latent factor (density) models. Such a connection has a number of useful benefits and implications for our machine learning practice.

Continue reading "A Statistical View of Deep Learning (II): Auto-encoders and Free Energy"

Variational Inference: Tricks of the Trade

The NIPS 2014 Workshop on Advances in Variational Inference was abuzz with new methods and ideas for scalable approximate inference. The concluding event of the workshop was a lively debate with David Blei, Neil Lawrence, Zoubin Ghahramani, Shinichi Nakajima and Matthias Seeger on the history, trends and open questions in variational inference. One of the questions posed to our panel and audience was: 'what are your variational inference tricks-of-the-trade?'

My current best-practice at present includes: stochastic approximation, Monte Carlo estimation, amortised inference and powerful software tools. But this is a though-provoking question that has has motivated me think in some more detail through my current variational inference tricks-of-the-trade, which are:
Continue reading "Variational Inference: Tricks of the Trade"