Meena Kotecha has been invited by Dr Michel d.S. Mesquita to co-present a workshop with him as part of the RSS 2020 online programme entitled “The art of blended and virtual learning in statistics and data science – insights from coronavirus quarantine”. Michel is a Founder and Independent Researcher at M2Lab.org, Centre for Statistical and Data Science Research (CSDS), Stord 5411, Norway. Michel has seen Meena's work on blended learning in statistics through her presentations at the past RSS annual conferences.

Zoom workshop session details:

Date: Tuesday 8 September 2020

Time: 1pm - 2.20pm

Registration: https://rss.org.uk/training-events/events/rss-2020-online-conference/

Order of Speakers:

1.       Morgan Yarker

2.       Meena Kotecha

3.       Michel Mesquita

4.       Matthew Benus 

Type of Presentations: Workshop style with potential audience participation. Live speaker session.

We are delighted to announce that Dr Moez Draief has been given a visiting appointment as a Visiting Professor in Practice for three years from 7 September. Dr Moez Draief is currently the Global Chief Scientist and VP of Data Science and Engineering at Capgemini, a multinational consulting company. In this role he works with clients in public and private sector organisations on their data science and data engineering delivery projects. Previously, he was the Chief Scientist for Artificial Intelligence at Noah’s Ark Lab of Huawei Technologies.

Before moving to industry in 2016, he worked as an academic for a number of years, including a position at Cambridge University as a Marie Curie Research Fellow and a position at Imperial College London as an Associate Professor. Furthermore, in the past two years, he has continued to offer MSc courses in the areas of online statistical learning (at ENSAE Paris) and Bayesian optimisation (at Université Paris Sciences et Lettres). He has continued being actively involved in research and having papers published in top-ranked journals and conference proceedings, including papers at NeurIPS, a premier international conference in machine learning area.

Welcome to the Department of Statistics, Moez!

A conference is taking place in Southampton on 8-10 July 2021 in honour of Chris Skinner and Fred Smith who both worked in the Department of Statistics and who both sadly passed away in the past year.  

There is currently a call for abstracts (deadline is 1 October) and further details of the conference will follow shortly. The conference will be based around the themes that interested Fred and Chris during their research careers.

Please see more information here. 

We are very excited to announce that our former PhD student Haziq Jamil, who graduated in September 2018, received an honourable mention for the prestigious Zellner Award! We are very proud. Have a read about the award details. 

Haziq worked on his PhD from 2014 to 2018 in the Department of Statistics at LSE. The title of his thesis was "Regression Modelling Using Priors Depending on Fisher Information Covariance Kernels (I-priors)."

Haziq is now an Assistant Professor in Statistics at Universiti Brunei Darussalam.

Well done for this achievement, Haziq! 

Bea Acciaio has sadly decided to leave the Department of Statistics for pastures new and has taken up a Professorship in the Department of Mathematics at ETH Zurich. We wish Bea every success in her new role and will be missed by everyone here in the department! Good luck on your new adventure, Bea! 

We would like to take this opportunity to congratulate Professor Pauline Barrieu on her new role as Academic Director in the School!  This is a two-year role to provide additional academic leadership as the School adapts its educational provision in light of the impact of COVID-19. She will be working closely with Professor Dilly Fung, Professor Julia Black, leaders of academic departments and service leads on the Curriculum Shift 2020.  Well done, Pauline! 

Dr Luciano Campi has sadly decided to leave the Department of Statistics for pastures new and has taken up a Professorship in the Mathematics Department at the University of Milan. We wish him every success in his new role and he will be missed by everyone here in the department! Good luck on your new adventure, Luciano! 

We are proud to announce that Phil Chan and Anastasia Kakou have won LSE Class Teacher awards! Many congratulations to them and all of their hard work which reflects their engagement, teaching innovation, excellent student feedback and dedication to the Department!

We are proud to highlight our own research in the Department of Statistics on the response to coronavirus which includes papers, summaries, opinions, trend estimates and blog posts. We will be updating the page regularly, but so far we have some fantastic contributions from Professor Clifford Lam, Professor Jouni Kuha and Professor Piotr Fryzlewicz. 

We are pleased to announce that MSc Quantitative Methods for Risk Management student Zhiyang Zhao and BSc Actuarial Science student Chenxing Zu in the Department of Statistics achieved great success in the Digital Skills Lab Python Coding Challenge, under very difficult circumstances. The challenge was organised by the Digital Skills Lab in partnership with LSE Careers and Global Technology Consultancy, Infosys. 

The challenge centred around financial analysis was set by Dr Antonio Mele (Econ) and Dr Srinivasa Raghavan (Chief Data Scientist, Data and Analytics) of Infosys, a large Indian IT consultancy, with the desire for students in their various teams, were to apply their coding skills to a real world problem. Aaron Trevitt from LSE Careers also worked with us on developing team work and employability. 

Due to the disruption of the Covid-19 outbreak the challenge had to be adapted to an online format at the last minute. This resulted in only four of the original nine teams completing their projects and submitting a final recorded presentation to compete for the four prizes. 

Despite the various difficulties encountered the challenge was a resounding success on several levels. Participants were able to demonstrate how they applied and extended their digital skills to a real-world problem, interweaving it with their own academic research on the topic set, and presenting their ideas clearly and effectively. 

The presentations can be viewed by LSE staff and students only using this link.

Professors Patrick Sturgis, Jon Jackson and Jouni Kuha have been successful in receiving a grant from the School of Public Policy's COVID19 fund to undertake research.

This survey research will focus on compliance with lockdown, the authority of the police, economic and health trade-offs and the impacts on employment, income, and education. The survey will be carried out via the Kantar Public Voice online probability panel, with which the Department has an existing collaborative agreement.

Patrick will also be taking part in the launch event of Kantar's Public Voice. This will present information on a powerful new random sample research tool and Patrick will give thoughts on the value of random sampling in survey research.

Matteo Barigozzi has decided to leave the Department of Statistics after 10 years with us and has taken up the role of Chair at the University of Bologna and has joined their Department of Economics.  We wish him every success in his new role and he will be missed by everyone here in the department! Good luck on your new adventure, Matteo! 

Congratulations to everyone who took part in the 2020 Industry Practitioners Challenge! On Wednesday 11 March, two teams of undergraduate students from the Department of Statistics presented their analysis on real life issues faced by Aviva. 

Boris Genot, Senior Quant Manager in Risk Analytics at Aviva, challenged students to analyse the interdependencies of risks in the internal model used by the Analytics and Capital Modelling Team. This had a direct impact in allocating capital and reporting to the regulator under the Solvency Capital Requirement.

Over the course of five weeks, students investigated four tasks:
1. To identify the optimal length for the time series to calculate stable correlations
2. To identify the optimal frequency for the observations in order to calculate stable correlations
3. To establish a method to annualise the correlations calculated with daily or monthly observations
4. To develop a robust model able to forecast reliable future correlations.

To tackle these tasks, teams used statistical criteria to decide which risk factors should be included in the model, cleaned the data, and implemented a Dynamic Conditional Correlation (DCC) model to estimate and forecast the risks (using computing languages Python or R). While both teams delivered strong presentations, one was chosen by Mr Genot to be overall winner. 

The Industry Practitioners Challenge is highly valued by students, with one student describing it as “great experience working with Aviva”. It’s organised annually by Dr Gelly Mitrodima, Assistant Professorial Lecturer in the LSE Department of Statistics. The Department collaborates with leading industry partners to provide an opportunity for students to work on competitive projects focusing on real issues faced by companies. 

Through the challenge students gain invaluable experience of working with industry experts and academic staff, as well as developing their communication and problem solving skills. Businesses (such as previous partners Aviva, JP Morgan, UBS and QBE) form strong links with LSE by helping nurture talent, while also benefitting from academics and students helping investigate problems.

Interested in working with undergraduate or postgraduate students from the LSE Department of Statistics? Speak to us about the 2021 Industry Practitioners Challenge today.

It is with great pleasure to let you know that the School Promotions Committee has decided that Dr Beatrice Acciaio and Dr Clifford Lam should be promoted to Professor from 1st August 2020! Congratulations to both of them on this achievement!

It is with great sadness that we share the news of Professor Chris Skinner's death. He passed away on 21st February 2020. Please read more about his fantastic accomplishments in a wonderful article written by his fellow colleague Jouni Kuha.

Professor Milan Vojnovic has received the Facebook’s Systems for Machine Learning Award for his research exploring how to optimise scheduling of machine learning jobs in data processing platforms. 

The Facebook’s Systems for Machine Learning request for proposals solicited impactful solutions in the areas of ‘scalable, elastic and reliable distributed machine learning and inference’ and other topics at the intersection of computer systems and machine learning. The call received 167 proposals from more than 100 universities in 26 countries around the world out of which 10 proposals have been selected for funding. The list of award winners and more information about the request for proposals is available here.

The aim of the research project is to investigate how should machine learning jobs, with uncertain, heterogenous processing requirements and complex delay costs, be scheduled in data processing platforms. This question falls in the core of systems for machine learning research area, aiming to satisfy metrics beyond predictive accuracy. This is a very challenging problem as machine learning workloads in modern data processing platforms (e.g. platforms for training machine learning models or content review in social media platforms) exhibit complex and uncertain delay costs which violate the assumptions of existing theory and practical solutions. 

The research will explore how to design novel job scheduling policies to enable efficient processing of machine learning workloads. This will require advancing the state of the art in online optimisation methods and the underlying theory, combining optimisation with statistical inference and learning in innovative ways. The project also aims to evaluate performance of different scheduling policies through experimentation, using existing solutions as benchmarks as well as policies learnt by using reinforcement learning algorithms. The project is expected to contribute new conceptual solutions that may lead to the design of next-generation system architectures.

Huge congratulations to Milan!

We would like to welcome to the Department of Statistics two brand new members of staff! Will Garrison is our new External Engagement Officer which is a brand new role in the department and also Shanice Kudita who is in the Office Co-ordinator position on a temporary basis. We are very excited for them to be working with us!

The Department of Statistics is sad to say that Kayleigh Brewer, who was our fantastic Office Co-ordinator for two years, has left LSE and has taken up a position as Communications Officer at Imperial College London. She has been amazing at supporting the Department and we congratulate Kayleigh on her new venture and she will be missed! Good luck!