Publications, Machine Learning Group, Department of Engineering, Cambridge

topics:
[ GPs | Clustering | Graphical Models | MCMC | Semi-Supervised | Non-Parametric | Approximations | Bioinformatics | Information Retreival | RL and Control | Time Series | Network Modelling | Active Learning | Neuroscience | Signal Processing | Machine Vision | Machine Hearing | NLP | Deep Learning | Review ]
current group:
[ Balog | Bauer | Bui | Dziugaite | Ge | Ghahramani | Gu | Hernández-Lobato | Kilbertus | Kok | Li | Matthews | Navarro | Peharz | Rasmussen | Rojas-Carulla | Rowland | Ścibior | Shah | Steinrücken | Rich Turner | Weller ]
former members:
[ Borgwardt | Bratières | Calliess | Chen | Cunningham | Davies | Deisenroth | Duvenaud | Eaton | Frellsen | Frigola | Van Gael | Gal | Heaukulani | Heller | Hoffman | Houlsby | Huszár | Knowles | Lacoste-Julien | Lloyd | Lopez-Paz | McAllister | McHutchon | Mohamed | Orbanz | Ortega | Palla | Quadrianto | Roy | Saatçi | Tobar | Ryan Turner | Snelson | van der Wilk | Williamson | Wilson ]
by year:
[ 2018 | 2017 | 2016 | 2015 | 2014 | 2013 | 2012 | 2011 | 2010 | 2009 | 2008 | 2007 | 2006 | 2005 | 2004 | 2003 | 2002 | 2001 | past millennia ]
GP

Gaussian Processes and Kernel Methods

Gaussian processes are non-parametric distributions useful for doing Bayesian inference and learning on unknown functions. They can be used for non-linear regression, time-series modelling, classification, and many other problems.


Clustering

Clustering

Clustering algorithms are unsupervised methods for finding groups of similar points in data. They are closely related to statistical mixture models.


Graphical Models

Graphical Models

Graphical models are a graphical representation of the conditional independence relations among a set of variables. The graph is useful both as an intuitive representation of how the variables are related, and as a tool for defining efficient message passing algorithms for probabilistic inference.


Monte Carlo

Monte Carlo Methods

Markov chain Monte Carlo (MCMC) methods use sampling to approximate high dimensional integrals and intractable sums. MCMC methods are widely used in many areas of science, applied mathematics and engineering. They are an indispensable approximate inference tool for Bayesian statistics and machine learning.


Semi-Supervised

Semi-Supervised Learning

Often, it is easy and cheap to obtain large amounts of unlabelled data (e.g. images, text documents), while it is hard or expensive to obtain labelled data. Semi-supervised learning methods attempt to use the unlabelled data to improve the performance on supervised learning tasks, such as classification.


Non-Parametric

Non-parametric Bayesian Learning

Non-parametric models are very flexible statistical models in which the complexity of the model grows with the amount of observed data. While traditional parametric models make strong assumptions about how the data was generated, non-parametric models try to make weaker assumptions and let the data "speak for itself". Many non-parametric models can be seen as infinite limits of finite parametric models, and an important family of non-parametric models are derived from Dirichlet processes. See also Gaussian Processes.


Approximations

Approximate Inference

For all but the simplest statistical models, exact learning and inference are computationally intractable. Approximate inference methods make it possible to learn realistic models from large data sets. Generally, approximate inference methods trade off computation time for accuracy. Some of the major classes of approximate inference methods include Markov chain Monte Carlo methods, variational methods and related algorithms such as Expectation Propagation.


Bioinformatics

Bioinformatics

Recent advances in biology have allowed us to collect vast amounts of genetic, proteomic and biomedical data. While this data offers the potential to help us understand the building blocks of life, and to revolutionise medicine, analysing and understanding it poses immense computational and statistical challenges. Our work in Bionformatics includes modelling protein secondary and tertiary structure, analysis of gene microarray data, protein-protein interactions, and biomarker discovery.


Information Retrieval

Information Retrieval

Information retrieval concerns develping systems that find material from within a large unstructured collection (e.g. the internet) that satisfy the user's need. The best example of such systems are web search engines, such as Google, but there are many other specialized applications of information retrieval (such as collaborative filtering and recommender systems). Information retrieval can be thought of as an inference problem: given the user's query, what are the relevant items in the data collection?


Reinforcement Learning

Reinforcement Learning and Control

We are interested in understanding the human sensory motor system from a mathematical, computational and engineering point of view. To do this, we need to use concepts from control theory, optimization, machine learning and statistics, as well as experimental methods based on human psychophysics and virtual reality. These formal tools are also useful for advancing robotics and decision theory.


Time Series

Time Series Models

Modelling time series and sequential data is an essential part of many different areas of science and engineering, including for example, signal processing and control, bioinformatics, speech recognition, econometrics and finance. Using basic building blocks such as hidden Markov models, linear Gaussian state-space models, and Bayesian networks, it is possible to develop sophisticated time series models for real world data. However learning (parameter inference / system identification) becomes computationally challenging for such sophisticated models.


Network Modelling

Network Modelling

Active Learning

Active Learning

Neuroscience

Neuroscience

Signal Processing

Signal Processing

Machine Vision

Machine Vision

Machine Hearning

Machine Hearing

Natural Language Processing

Natural Language Processing

Deep Learning

Deep Learning

Reviews

Review Articles and Tutorials