Machine Learning Seminar
Thursday, 1 June, 2pm, ELG10
Jakob Foerster (University of Oxford)
Title: Counterfactual Multi-Agent Policy Gradients
Abstract: Cooperative multi-agent systems can be naturally used to model many real world problems, such as network packet routing or the coordination of autonomous vehicles. There is a great need for new reinforcement learning methods that can efficiently learn decentralised policies for such systems. To this end, we propose a new multi-agent actor-critic method called counterfactual multi-agent (COMA) policy gradients. COMA uses a centralised critic to estimate the Q-function and decentralised actors to optimise the agents’ policies. In addition, to address the challenges of multi-agent credit assignment, it uses a counterfactual baseline that marginalises out a single agent’s action, while keeping the other agents’ actions fixed. COMA also uses a critic representation that allows the counterfactual base- line to be computed efficiently in a single forward pass. We evaluate COMA in the testbed of StarCraft unit micromanagement, using a decentralised variant with significant partial observability. COMA significantly improves average performance over other multi-agent actor-critic methods in this setting, and the best performing agents are competitive with state-of-the-art centralised controllers that get access to the full state.
When: 15 May 2017, 1pm
Abstract: Appropriate feature engineering, the process of combining and transforming raw features, can contribute significantly to improving the performance of a supervised learning task. Because of the combinatorial explosion of possible engineered features, very often successful feature engineering in a corporate environment requires domain knowledge. However, with the increasing complexity of data and numbers of sources of data, applying human insight to the feature engineering process becomes difficult.
In this presentation, I introduce a novel, fast, and hopefully elegant algorithm where the complexity of the process of feature engineering is only connected to the number of features relevant to modelling the target variable, instead of all the features in a dataset. When the number of relevant features is much smaller than the total number of features, this amounts to increasing the efficiency to almost constant time. This method draws in insights from previous work linking machine learning and information theory. I then present the results of tests of the algorithm, which show that the engineering method I have developed is indeed effective in creating a feature that improves the performance of both classification and regression algorithms when the engineered feature is included along with the rest of the features. In order to reach statistically valid conclusions, it is necessary to test the algorithm on large numbers of appropriate datasets. Therefore I also introduce a method by which to generate synthetic datasets with desirable characteristics on which to test machine learning algorithms.
At this stage, the work is a proof of concept of the method created. Future work would include creating more generalised methods and coding these into Python packages for use by the data science community.
Hosted and organised by the Students’ Union, the Learning Enhancement Awards allow City students to recognise staff for their support (https://www.culsu.co.uk/student-voice/lea/)
Member of the Machine Learning Research Centre, Greg Slabaugh received the Research Student Supervision award 2017, an award that covers all Schools at City, University of London.
Machine Learning seminar
Prof Duncan Gillies, Imperial College London
Monday, 20 March 2017, 2pm. College Building, room AG22
Title: Visual Cognition in Face Recognition
(work by C. E. Thomaz, V. Amaral, G. A. Giraldi, D. F. Gillies and D. Rueckert)
Abstract: Recently new methods have been emerging in the field of image recognition. In particular, deep neural networks have reached very high levels of accuracy out performing other learning paradigms on large image data bases However, although deep networks are biologically inspired they tell us very little about how human visual perception works. They can correctly classify a picture, but they cannot give us any interpretation of that picture. Human vision has been found to operate quite differently. When viewing a picture we tend to concentrate our gaze at a small number of key points rather than looking at it globally. Our efforts are directed towards interpreting rather than classifying. An interesting question is “is it possible to characterise and make use of the mechanisms of human visual perception?”
The talk will focus on face recognition, and introduce some recent work carried out as a joint project between Centro Universitario FEI and Imperial College on incorporating human visual cognition with existing face recognition algorithms. It will give a brief overview of the history of face recognition including Sirovich and Kirby’s work on face spaces using PCA and Yarbus’ eye tracking experiment on human perception. It will then go on to describe some experiments where human fixation points have been used to spatially weight the PCA face feature space leading to improved accuracy of recognition.
The 12th edition of the Workshop on Neural-Symbolic Learning and Reasoning will take place at City, University of London, on 17 and 18 July 2017.
The workshop studies the combination of well-founded symbolic AI with robust neural computing to help tackle data-driven challenges in many areas of application, from health to finance, transport and global business.
Information about paper submission deadlines, how to register, and the new industry track can be found at http://www.neural-symbolic.org/NeSy17
19 Dec 2016, AG05, 2pm-5pm
Logic Tensor Networks, Knowledge Extraction and Applications
Ivan Donadello (FBK, Trento)
Adriana Danilakova (City, London)
Luciano Serafini (FBK, Trento)
Edjard Mota (UFAM, Manaus and City, London)
Simon Odense (City, London)
Dan Philps (Mondrian and City, London)
All are welcome!
ps. CoCo@NIPS’16 was a great workshop: http://daselab.cs.wright.edu/nesy/CoCo2016/
Who: Dr Eric Humphrey
When: Monday 5th Dec 2016, 16:00-17:00
Where: AG08 College Building, City, University of London
Title: Open-MIC – The Open Music Instrument Classification Challenge
Abstract: The Open-MIC initiative is a community-driven experiment to benchmarking content-based MIR algorithms in a transparent and sustainable way. Eric will present the inspiration behind Open-MIC, the approach to building an open music dataset, and the engineering effort behind making the project work. Open-MIC is focused on music, but these topics that are widely applicable to general machine learning and signal processing evaluation.
About the speaker: Eric Humphrey is a Senior Machine Learning Researcher at Spotify, focusing on algorithms and technology related to audio experience. He obtained a BSEE at Syracuse University in 2007, an MSc in Music Engineering Technology (U of Miami) in 2009 and a PhD in Music Technology at NYU in 2015. Eric is also a multi-instrumentalist, has been a visiting lecturer at the University of Miami, worked as an independent contractor roles for several audio technology companies, spent a summer at Google doing really fun things he can’t really talk about, and currently serves as the secretary to the International Society for Music Information Retrieval.
When: Tue, 6th Dec 2016, 1pm
Where: ELG04, City, University of London
Speaker: Dirk Nachbar (Google, marketing analytics)
Title: Analytical Problems in Marketing Science
Abstract: Like in many fields data is growing exponentially in marketing. I will describe four applications from the real world that use a lot of data and require modelling or data driven approaches: loyalty card data, online attribution problem, real time bidding/buying, and card linked marketing. The talk will introduce challenges and potential solutions in marketing and will describe the different players in the marketing ecosystem.
Muhammad Asad and S.M. Masadur Al-Arif who are pursuing PhD degrees in the Computer Vision Group within the Research Centre for Machine Learning have received Best Paper Awards for research presented in two leading conferences this year.
In June 2016, Muhammad received the Best Paper Award at the 2nd Workshop on Observing and Understanding Hands in Action, part of the 29th IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, USA, for the paper “Learning Marginalization through Regression for Hand Orientation Inference”, co-authored by Dr Gregory Slabaugh.
The paper presented a novel machine learning model that enables a computer to understand and describe the orientation of a human hand by looking at colour images of it.
In September 2016, Arif received the best paper award at the Fourth MICCAI Workshop on Computational Methods and Clinical Applications for Spine Imaging, part of the 19th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), Athens, Greece, for the paper “Global Localization and Orientation of the Cervical Spine in X-ray Imaging”, co-authored by M. Gundry, K. Knapp, and G. Slabaugh.
The paper presented a machine learning framework capable of teaching the computer to locate the spinal column in X-ray images. It learns what the spinal column looks like from expert annotated example X-ray images and is able to locate the spine on unseen images.
For more information, please visit:
Date & Time: Monday 28 Nov at 11 am
Venue: A110 (College Building)
TITLE: Sub-chronic recordings for myoelectric control of prostheses: paving the way for a better understanding of the language of the brain
ABSTRACT: Every day somewhere globally, at least one person undergoes amputation of part or entire upper limb. An estimated prevalence of 11.6 per 100,000 has been reported in Norway. The social and economic burden of limb deficiency has led to development of artificial means to replace missing functions. Among the available solutions, the most advanced in clinics are those based on active control using the electric activities (electromyographic, EMG, signals) produced by the remnant muscles. The use of EMG to control prosthetic limbs dates back to more than 50 years. Progress has been made in this period, mainly in the laboratory settings. However, even though multi-dexterous prosthetic hands (e.g. DEKA Hand) are now available, the control has remained limited to the original solutions in the clinics. The use of machine learning (ML), rather than direct control, has attempted to advance the control possibilities of the users but it is limited by unsatisfactory robustness to non-stationarities (e.g. changes in electrode positions and skin-electrode interface). Robustness is the key characteristics of any clinical solution. Very advanced control systems that allow a substantial functional benefit for short-term, laboratory tests, cannot be translated into clinical solutions if their performance worsens over time. This presentation gives an overview of our clinical and laboratory understanding of the language of the brain (“Brainish”) in myoelectric control of prostheses. Furthermore, preliminary results of sub-chronic experiments (seven days) are presented together with future perspectives.
Ernest Nlandu Kamavuako received the Master and Ph.D. degrees in biomedical engineering from Aalborg University, Aalborg, Denmark, in 2006 and 2010. Since 2014, he has been an Associate Professor at the Department of Health Science and Technology, Aalborg University, Denmark with excellent teaching and supervision skills. In 2015, he was named teacher of the year by the students of the study board for health technology and Sport science. From 2007 to 2008, he was a Research Scholar in the Biomedical Department, IUPUI, Indianapolis, USA. From 2012 to 2013, he was a Postdoctoral Fellow at the Institute of Biomedical Engineering, University of New Brunswick, Canada. He has a good publication record with main research interests related to the use of invasive recordings in the control of upper limb prostheses. Other research interests include muscle recovery functions following electrical stimulation, applied signal processing and the application of near-infrared spectroscopy and EEG for brain–computer interfaces.