Title: Understanding Big Data (and not die trying)
Speaker: Dr Serge Plata (Head of Analytics, Home Retail)
King William KW303, University of Greenwich, Old Royal Naval College, London, SE10 9LS
There are so many terms in the analytics industry today, and no one really understands them well: Data science, data analysis (or is it analytics?), data architecture, web analytics, business intelligence, and management information; there are also many platforms from relational databases to Hadoop clusters, and programming languages like python, .NET, C# and F# (sounds like music!); not to mention all the statistical packages like (R, SPSS and STATA) and front end tools (like Pentaho, D3.js or Tableau). But what does this mean? Where do all of these elements fit in the data and analytics profession?
In this talk I will explain what they are, how they work and how they interact with each other, but most importantly how businesses use them to take advantage of all these mathematical tools and make data-driven decisions. I will give real life examples on these tools and methods like machine learning, data mining and mathematical modelling and also how platforms are used.
Dr. Plata did his doctoral research at Imperial College London before taking up a post as a research fellow at the University of Exeter. Prior to that, he did a degree and masters in mathematics, specializing firstly on differential and algebraic topology and moving then into spectral theory, homeomorphic dynamics and ergodic theory which classically fall into the applied fields like optimisation, game theory, and machine learning. Among other publications, he wrote a book under Peter Lang Publishers titled "Visions of Applied Mathematics"
In terms of mathematical applications, Dr. Plata has extensive experience building, and developing analytics programmes as well as leading data projects and data science teams from FTSE100 companies to technology SMEs. He has worked mainly in the retail space including digital and mobile, pioneering on behaviour analytics, machine learning and big data.
He is a fellow of the IMA and he currently heads the 'data science and advanced analytics' team at Home Retail.
Date and time: Wednesday 4th May 2016 (6.30-7:30pm)
Abstract: P-values have become a staple measure used in hypothesis testing, an interesting application of which is the field of forensic statistics. Advances in technology and forensic data collection allow statistics to be applied to the law - specifically the probabilistic assessment of a defendant's guilt based on data evidence. This talk considers the use of p-values to determine the likelihood of guilt while attempting to appease the different 'schools' of hypothesis testing which advocate alternative approaches to testing - the battle between the Frequentists and the Bayesians.
Turbulent plumes are free-shear flows driven by steady (timeindependent) or unsteady (time dependent) sources of momentum and buoyancy.
Relatively simple integral models of plumes are useful in practical applications such as pollution modelling, firemodelling and the natural ventilation of buildings. The models are obtained by integrating the partial differential equations of motion (the Navier-Stokes equations) and invoking various assumptions. In this talk we will relax these assumptions anddiscuss plume theory from a more general perspective. In particular, we will derive generalised unsteady plume equations and demonstrate that they have mathematical properties andphysical implications that have previously been overlooked. The talk will provide an overview of classical steady plume theory before considering planar and axisymmetric unsteady plumes.
Wednesday 21st October 2015, 6:30pm
QA065, Queen Anne Court, University of Greenwich. (For travel information and a campus map, go to: http://www2.gre.ac.uk/about/travel/greenwich)
Forecasting an Ebola Epidemic using Mathematical models – Dr Sebastian Funk (London School of Hygiene and Tropical Medicine)
Abstract to follow
I'm sure we've all had a bit of a groan when we receive our car insurance renewal premiums; typical questions which arise are, "How can the premium be so high when I’ve never had an accident?", "Why has my premium gone up when I have one more year's no claims discount?", and "Why won’t my current insurer price-match the cheapest quote I can obtain?". It's not always as cut and dried as you might think and this talk will explain what is going on.
We will examine the theoretical statistical basis of risk (multivariate probability distributions), EU regulations concerning theoretical "probabilities of ruin", theoretical "customer lifetime value" models as well as seeing the theories in action with some real-time examples from price comparison websites. We will also look at forward pricing, statistical quirks in the data and some pricing anomalies, all explainable from the theoretical considerations.
So, are you paying too much for your car insurance? Come and listen to the evidence for and against, given by an experienced and independent actuary (and car insurance payer!) deeply involved in insurer pricing, solvency and regulatory management of insurers - and draw your own conclusions!
Last year (2013) Teach First recruited 230 maths teachers. Whilst the majority of Teach First ambassadors remain in the classroom, our Ambassadors are also playing a key role in government, civil service and the wider policy community, as well as driving engagement with the corporate world through leadership roles in business. The session will provide background to the Teach First organization through a discussion of the unique partnership with industry and Teach First’s own mission to end inequality in education by building a community of exceptional leaders who create change within classrooms, schools and across society. The session will be presented by Piers Saunders, the National Lead for Mathematics within Teach First, an experienced mathematics teacher as well as a lecturer in Mathematics education at the Institute of Education.
Tuesday 24th June 2014 (6.00pm for 6.30pm start) 'Fail safe or fail dangerous?', Dr Ahmer Wadee (Imperial)
Abstract: Structural engineering has a rich history with its landmark monuments and numerous personalities. As a discipline, it relies fundamentally on nonlinear mathematics, which can reveal whether structures are vulnerable to catastrophically dangerous collapse (or failures) or those that can be contained. Structural failure is inherently a nonlinear process with the nonlinearities grouped into two broad categories: material and geometric. Material nonlinearities arise either intrinsically from the material response (e.g. concrete, aluminium or timber) or when materials that obey Hooke's law (e.g. steel) are overstressed and permanent damage or fracture occurs and sudden changes in material behaviour are observed.
We will focus our attention to geometric nonlinearities, which are generated in structures experiencing large deflections. These may rapidly occur when the phenomenon known as buckling, when structures change from stable to unstable equilibrium, is triggered. Buckling is most likely in components that are made from slender elements that are wholly in, or in-part, compression such as columns and beams. After buckling is triggered a small change in load may lead to disproportionately large deformations and structural collapse. We shall see that nonlinear mathematics provides key tools that allows engineers to assess robustness and make informed decisions about whether potential failure modes are catastrophic or otherwise. Recent developments regarding the process known as "cellular buckling" will also be outlined; under certain
circumstances, imaginative use of that buckling process can even be exploited to enhance safety.
Pigs didn't fly, but swine flu by Ken Eames and Ellen-Brooks Pollock (Ken is a Lecturer in the London School of Hygiene and Tropical Medicine, and Ellen is a Research Fellow at the University of Cambridge)
6:30 pm for 7:00pm start Tuesday 4 February 2014
Venue: Imperial College London, South Kensington Campus. Room 201. Skempton Building. London SW7 2AZ (map of campus)
Nearest Tube: South Kensington – take the tunnel to the end, turn left out of the exit and left again.
Abstract: Traditional symmetric or secret key cryptography has historically been the province of government, diplomatic and military users with a requirement for confidentiality. More recently asymmetric or public key cryptography has been developed and taken up to meet the requirements of life in cyberspace. Private individuals are now using it, often unknowingly, to achieve confidentiality and authentication. In this talk we describe how some of these requirements are met.
Maths is Everywhere by Professor John Barrow
18:30 - 19:30, Tuesday 23 October 2012
Venue: UCL, Anatomy Building, G29 J Z Young Lecture Theatre, Gower Street [location
All welcome. Please contact Noel-Ann Bradshaw (firstname.lastname@example.org) to book a place at this event.
Mathematics and the Cutty Sark - Saving a National Maritime Treasure by Professor Chris Bailey (Greenwich)
18:00, Thursday 13 October 2011
Venue: Queen Anne 180, University of Greenwich, Old Royal Naval College, London SE10 9LS
There will be no charge for this meeting. Please note that there is no parking on campus.