The Manchester MLO Group conducts world-leading research into a wide
range of techniques and applications of machine learning, optimization,
data mining, probabilistic modelling, pattern recognition and machine
perception. The group spans the field from new theoretical developments
to large applications, and is currently supported by a
number of research bodies, including EPSRC, BBSRC, and several
...hide older news items...
Congratulations to MLO members Ming-Jie Zhao, Narayanan Edakunni, Adam Pocock and Gavin Brown,
on their new JMLR paper Beyond
Fano's inequality: bounds on the optimal
F-score, balanced error rate, and cost-sensitive risk using conditional entropy and their
Congratulations to MLO members Joe Mellor and Jon Shapiro on their accepted AISTATS paper
Thompson Sampling in Switching Environments with
Bayesian Online Change
Congratulations to MLO members Hassan Bashir and Richard Neville on two new papers now available on IEEExplore.
A Hybrid Evolutionary Computation Algorithm for
Global Optimization, and Convergence measurement in evolutionary
computation using Price's theorem, which were published in the IEEE Conference on Evolutionary Computation in June.
Congratulations to MLO member Peter Glaus on the publication of a new article
in the Journal of Bioinformatics, titled
Identifying differentially expressed transcripts from RNA-seq data with biological variation.
Gavin Brown delivered a talk at ICML on the recent JMLR
Gavin Brown was interviewed on BBC Radio Manchester regarding the REUNITE project.
The University has also issued a press release, and
the Manchester Evening News also ran
the story. UPDATE: The BBC World Service's flagship
technology programme "Click" also featured the project. The podcast
Congratulations to MLO members
Gavin Brown and (APT member)
on acceptance of their paper to AISTATS 2012, titled
'Informative Priors for Markov Blanket Discovery'.
Joshua Knowles went to the Smart Cities conference in Westminster
to learn about opportunities for ML and Optimization in the future of
waste management, transportation, communications...and a lot more, to
help our cities become greener and more sustainable.
Congratulations to Gavin Brown, Adam Pocock, Ming-Jie Zhao, and Mikel Lujan on acceptance of their
JMLR paper 'Conditional Likelihood Maximisation: A
Unifying Framework for Mutual Information Feature Selection'. Check back for a pre-print in a
We are pleased to report the successful completion of the REUNITE project.
Congrats to MLO members Richard Allmendinger and Joshua Knowles for their paper "Evolutionary Search in Lethal Environments" at
ECTA 2011, which was nominated for the Best Paper Award !
Congrats to MLO members Ahmad Salman and Ke Chen, for two recent publications
accepted in IEEE TNN and NIPS 2011 !!
Chen K. & Salman A., Learning speaker-specific characteristics with a deep neural architecture.
IEEE Transactions on Neural Networks and Learning Systems, vol. 23, 2012.
Chen K. & Salman A., Extracting speaker-specific information with a regularized Siamese deep network.
Advances in Neural Information Processing Systems 25 (NIPS'11), MIT Press, 2011.
MLO group members Xiaojun Zeng and Ke Chen are hosting the UKCI 2011
workshop here at Manchester, Sept 7-9. The keynote lectures will be given by Prof Steve Furber and Prof Sheng
New paper published by MLO members Joshua Knowles and Richard Allmendinger.
B.G. Small, B.W. McColl, R. Allmendinger, J. Pahle, G. Lopez-Castejon,
N.J. Rothwell, J. Knowles, P. Mendes, D. Brough, and D.B. Kell (2011):
Efficient discovery of anti-inflammatory small molecule combinations using
evolutionary computing . Nature Chemical Biology, (In Press).
Nara Edakunni is joining the group as an RA on the iTLS project - welcome Nara!
Paper on "Boosting as a Product of Experts" (Edakunni, Brown, Kovacs) appearing at UAI 2011.
Paper on "Policy Learning in Resource-Constrained Optimization." (Allmendinger, Knowles) appearing at GECCO 2011.
July 2011: The group has invested in a high-performance multi-core
machine for compute intensive applications of machine learning.
The new Dell PowerEdge 610 server has 12 Intel Xeon 3.46ghz cores,
with 32gb RAM. This resource will support our current
research into various avenues, including
large scale image analysis with deep neural network architectures
and multi-objective optimisation problems.
July 2011: AstraZeneca Research are sponsoring 5 MSc bursaries this year, to study machine learning technologies for drug analysis.
This is part of their predictive safety science initiative. MSc students Eli Marinoiu
and Ahmed Osman are part of the MLO group team awarded this project. Well done!
July 2011: Ke Chen will be presenting a
tutorial on learning in deep neural architectures at IJCNN in California.
July 2011: UG student Matt Leach has joined the group as a summer intern, studying hardware plausible implementations of machine learning
Joshua Knowles gave an invited talk, "Evolutionary algorithms for
multiobjective optimization: current trends" to the Decision Analysis
Special Interest Group (DASIG) of the Operations Research Society,
University of Portsmouth, June 9th 2011.
Gavin Brown delivered an invited
series of doctoral lectures on Feature Selection methods at the University of Cagliari, Sardinia.
Joshua Knowles ran a stand at the National Science and Engineering
Week Fair, University of Manchester, with 3rd year project
students, Osman Osman and James Whittaker. The fair ran for three days
and was visited by pupils from 58 local schools. The stand had two
interactive visual simulators of "Swarm Intelligence" algorithms.
Want to study with us for a PhD?
deep neural nets,