Bio
I have a BA in Experimental Psychology and Philosophy from the University of Oxford, and an MSc in Psychological Research Methods and Statistics from the University of Liverpool. I'm 25 years old and originally from London but have been living in Liverpool for over 2 years now, so I am really inconsistent when pronouncing "bath" and "grass".
Experimental Psychology, Behavioural Neuroscience, Neuroeconomics.
Research project title: Neural mechanisms of economic decision making during online purchasing
Supervisory team: Dr Andrej Stancak, Dr Olga Gorelkina, Dr Yiquan Gu, Dr Marco De Angelis, Dr Amogh Deshpande
Bio
I graduated from Queen's University Belfast in 2017 in Physics, and have been a student at the Institute for risk and uncertainty since October of 2017.
My work mainly revolves around Monte Carlo methods for particle transport, uncertainty propagation, and Bayesian analysis. My background in Physics was in atomic simulation, computation of the electronic structure of matter using Density Functional Theory.
Monte Carlo Simulation, Bayesian Analysis, Validation, Neutron Transport
Research project title: Quantification of Nuclear Data Uncertainy in Fusion Neutronics
Project description: Neutronics is the study of how neutrons propagate through matter: how they interact, what with, and their manipulation. It is a basic branch of fusion research. Fusion neutrons must be used in tritium breeding and will be an energy production mechanism through energy deposition blankets. Neutronics will also bring control over hazards like radiation exposure and waste production, key to the success of experimental reactors and future power stations. The Nuclear Data governing how neutrons interact with matter is the main ingredient in any neutronics calculation, yet is very uncertain. A lacking of experimental data for the vast majority of nuclear reactions has resulted in nuclear data being produced with an inherent uncertainty value. The impact of Nuclear Data uncertainty is exaggerated in fusion since the number of nuclides and materials used is more diverse than fission. This projects primary aim is to study the impact this uncertainty has on ITER scale neutronics problems. To reach this goal, efficient methods for uncertainty propagation in neutron transport codes will be developed, as current methods are too computationally intractable for ITER scale simulations in the majority of situations.
Supervisory team: Edoardo Patelli (Risk Insititute), Andrew Davis (CCFE)
ORCID: 0000-0002-1585-0900
Bio
My PhD research project is based on the multi-disciplinary EPSRC and ESRC Centre for Doctoral Training (CDT) on Quantification and Management of Risk & Uncertainty in Complex Systems & Environments, within the School of Law and Social Justice and the Institute for Risk and Uncertainty. The purpose of my research study is to map the methodological practices and decision support tools employed in assessing risks in risk management organisations.
I have joined the University of Liverpool–Institute for Risk and Uncertainty in September 2016, studying an MRes in decision making under risk and uncertainty. Prior to that, I studied an MSc in Risk Management at Glasgow Caledonian University, graduating in 2015 with distinction. I have also completed a BA in Social Sciences and Community Development from the University of Glasgow and hold a BSc in Economics. I am passionate and enthusiastic about risk management and its related context, and have published two papers in international refereed conferences.
Mathematics, Social Sciences
Research project title: Living and Future Tools for Risk Assessment: An Examination of the Possibilities for Fusion
Project description: The purpose of my PhD research is to map the methodological practices and decision support tools employed in assessing risks in risk management organisations. The study will involve analysing different methodologies and decision support tools that are used across a range of risk regulation agencies dedicated to assessing risks.
Supervisory team: Professor Gabe Mythen, Department of Sociology, Social Policy and Criminology Dr Hirbod Assa, Department of Mathematical Sciences Dr Beverley Bishop, Chief Social Researcher, Health and Safety Executive
ORCID: 0000-0002-2884-2467
Bio
Benjamin studied BSc Mathematics at Queen Mary, then MSc Mathematics at Liverpool. Support Everton and once won the North-West’s handwriting competition.
Mathematics and Management
Research project title: Food Supply Chain Risk Management: an Insurance Perspective.
Bio
I have been working in the industry for 20 years, most of the time in Oil & Gas. Opting to pursue a PhD and therefore stepping back in my industry career turned out to be necessary, as I had lots of questions related to Human Factors accumulated throughout those years in Safety Inspections and Accident Investigation. Here in the Risk Institute, I receive specialised support on how to find out the risks and uncertainties of my area of interest, using probabilistic and uncertainty quantification tools to systematise the state of the art and my ideas. My research is on Human Reliability in Engineering.
Engineering (Medical, Aerospace, Civil, Mechanical, Biological, Software, etc.), Psychology, Cognitive Science
Research project title: Modelling Human Reliability in the Design Phase
Project description: The aim of the research is to understand which factors impact the performance of engineers and decision makers during the desigh phase, using existing methods of Human Reliability Analysis (HRA).
However, there is a lack of data of factors for the desigh phase - as the focus of research is on data of the operational phase.
For this reason, this research also have a second objective running in parallel: to enhace the state-of-the-art of Human Reliability Analysis (HRA) in the operational phase, providing HRA community with Human Error Probabilities (HEP) based on data extracted from investigation reports of major accidents.
The hypothesis is that this kind of operational data is better than simulating data (from control room simulation) and expert judgement.
The probabilistic tool used is the Bayesian network.
Supervisory team: Edoardo Patelli, Michael Beer, Raphael Moura
ORCID: 0000-0002-9329-4110
Bio
PhD Student and Research Assistant at the Institute for Risk and Uncertainty and the Critical and Major Incident (CAMI) research groups.
Accelerated learning in the security, defence and emergency services. Specific interests involve autonomous evaluation of investigative interviewing performance and real-time decision-making in high-stress and high-stakes settings.
Research project title: Training & Learning Evaluation Frameworks: Monitoring Skills & Knowledge Audits
Project description: The aim of the research is to understand which factors impact the performance of engineers and decision makers during the desigh phase, using existing methods of Human Reliability Analysis (HRA).
However, there is a lack of data of factors for the desigh phase - as the focus of research is on data of the operational phase.
For this reason, this research also have a second objective running in parallel: to enhace the state-of-the-art of Human Reliability Analysis (HRA) in the operational phase, providing HRA community with Human Error Probabilities (HEP) based on data extracted from investigation reports of major accidents.
The hypothesis is that this kind of operational data is better than simulating data (from control room simulation) and expert judgement.
The probabilistic tool used is the Bayesian network.
Supervisory team: Prof. Laurence Alison, Prof. Simon Maskell, Dr. Michael Humann
ORCID: 0000-0002-9329-4110
Bio
I am a doctoral researcher working in the Institute for Risk and Uncertainty at the University of Liverpool. Broadly, my interests include seismic reliability assessments for structures and systems, reliability-based optimisation techniques and fragility analysis. More specifically, I am interested in the use of advanced probabilistic algorithms ( e.g. Artificial Neural Networks, Classification algorithms, Metamodels , Reliability analysis, Sensitivity analysis etc.) ,mostly coming from the machine learning field , to account for the uncertainty that affect the seismic reliability assuming different failure criteria. The same methodologies can be easily applied for the uncertainty quantification and risk management in a wide range of applications and my collaboration with the nuclear industry Areva GmbH is an example of that.
Seismic reliability analysis; reliability-based optimisation; fragility curves
Research project title: Seismic reliability assessment of structures and systems
Project description: The safety of civil structures and the protection of their human occupants represent goals of primary importance during the design stage. Events like earthquakes and extreme winds are the main causes that generate the need for dynamic protective measures of such structures. One way to prevent and mitigate the effect of dynamic loads on such structures is by means of hydraulic devices able to dissipate kinetic energy during an earthquake or extreme wind. The optimal design of viscous dampers under uncertain seismic excitations has represented one of the main research topic of my first PhD year. In addition, as result of a strong collaboration with the industrial partner AREVA GmbH, the research is also focused on the analysis of dynamic interactions between fuel assemblies in case of a seismic event. A first study estimates fragility curves of contiguous fuel assemblies in a nuclear reactor , analyzing the effects of dynamic impacts between spacer grids. Secondly, for a more general approach, a simplified structural system is analyzed to provide a dimensionless probabilistic demand model to estimate the risk connected to the pounding phenomena between adjacent systems. Finally, a further research topic that aim at defining a non-parametric efficient approach to evaluate conditional failure probabilities and seismic fragility curves is under development.
Supervisory team: Edoardo Patelli, Michael Beer, Andreas Rietbrock, Hector Jensen
ORCID: 0000-0002-9441-388X
Bio
Dominic Calleja has a background in aerospace engineering, nuclear engineering, and uncertainty quantification. Dominic has broad interests in the field of uncertainty quantification (UQ), and has experience in the application of UQ methods to industrial scale problems across many industries and sectors. Dominic was appointed academic manager, and impact follow at the Risk Institute in July 2018. He is passionate about the communication of risk, and particular interests in intuitive representations of uncertainty.
Nuclear Fusion, Machine Learning, Imprecise Probability, Efficient Numerical Methods, Sensitivity Analysis, Robust Optimisation
Research project title: Strategy for Sensitivity Analysis of DEMO first wall
Project description: His PhD is primarily focused on the plasma-wall interactions at the edge of thermonuclear fusion devices, focussing on the effect of uncertainties. In collaboration with the Culham Centre for Fusion Energy (CCFE), Dominic develops software tools for the assessment of uncertainty in fusion devices.
Dominic engages in the development of efficient algorithms for the treatment of uncertainty in black-box models, and novel approaches to the propagation of imprecise probabilistic structures.
Supervisory team: Edoardo Patelli
Bio
I have an MSc in geological and environmental hazards with a focus on flank instability in volcanic edifices. Subsequently I worked as international project manager in Munich promoting the use of ESA Sentinel satellite imagery. My current research involves combining satellite imagery with time lapse field data using the latest remote sensing techniques to assess the risks of icebergs for the public and industry in the Godthåbsfjord area, SW Greenland.
Glaciology, Remote Sensing, Python Programming, Data Science, Image Processing, Risk communication, Archaeology
Research project title: Iceberg Calving in Populated Fjords: Past, Present and Future Risks
Project description: This project aims to produce new scientifc insights into the calving behaviour and retreat of tidewater glaciers in Greenland, and identify possible risks of icebergs in populated in the Godthåbsfjord (Nuup Kangerlua) area, SW Greenland. The research will have a specifc focus on the glacier Narsap Sermia, due to its close proximity to Greenland's capital Nuuk, which is also home to the largest deep-water port in Greenland. The aims of the project are to investigate the past behaviour of Narsap Sermia in order to make more accurate predicitions about the future while also creating a tool to determine iceberg distribution in the fjord and communicate iceberg associated risks.
Supervisory team: Dr James M. Lea, Dr Clare Downham, Dr Jakob Abermann, Prof Douglas Mair
ORCID: 0000-0002-7895-1557
Bio
I am a social scientist with a background in International Development, with a particular interest in agriculture and livelihoods in Eastern and Southern Africa.
I hold a BA in International Development, and an MA in Environment, Development, and Policy; both from the University of Sussex. My dissertations looked at conservation through commercialisation of the argan forest, and climate vulnerability of smallholder farmers in Northern Uganda respectively.
Since moving to the University of Liverpool I completed my Mres in decision making under risk and uncertainty, with my dissertation looking at smallholder access to financial services in Gulu, Uganda, a topic relevant to my PhD.
I am currently based in Busia, Kenya for fieldwork, hosted by the International Livestock Research Institute's ZooLink (zoonosis in livestock in Kenya) project.
Rural livelihoods, agriculture, smallholder farming, international development
Research project title: Farming in Transition in East Africa: Financial Risk Taking and Agricultural Intensification
Project description: 'Farming in Transition in East Africa: Financial Risk-Taking and Agricultural Intensification' considers both financial and social risks that occur as a result of farmers attempting to intensify from subsistence to commercially oriented models of farming. The study will be using a mixed-method longitudinal approach in the form of a Financial Diary, adapted from the Financial Diary methodology developed by Bankable Frontier Associates (see financialdiaries.com). Methods that will be used include financial diary interviews (cash flow statements and balance sheets), questionnaires, ethnographic interviews, observation, photography, and possibly semi-structured interviews and focus groups. Therefore, this study will provide a wealth of human side data relating to intensification, the reason farmers choose to intensify their production, and the complexities of the impacts this has on the homesteads.
Supervisory team: Prof. Eric Fevre, Prof. Jude Robinson, Dr. Rob Christley
ORCID: 0000-0002-9269-5887
Bio
Elpida graduated from the National and Kapodistrian University of Athens in 2017 with a 5-year pharmacy degree. In September 2017, she joined the Centre for Doctoral Training in Quantification and Management of Risk & Uncertainty in Complex Systems & Environments at the University of Liverpool. Her research lies among drug safety, statistics and computer science, with her project being focused on developing novel algorithms for signal detection of drug-drug interactions. She is co-funded by EPSRC and AstraZeneca.
Signal detection, Pharmacovigilance, Bayesian Statistics
Research project title: Improved Methods for Drug-Drug Interaction Detection
Project description: Adverse drug–drug interactions (DDIs) constitute a major issue in clinical practice and are related to a large number of hospital admissions each year.
Since clinical trials for a drug are necessarily limited in their scale relative to the downstream use of the drug, the safety profile of new medicines cannot be fully determined and understood before they reach the market. Also, the increasing rates of polypharmacy in recent years make the problem even more complex. Thus, post-marketing surveillance is a vital stage in the lifecycle management of a medicine and reporting databases exist in order to collate information about patients who experience adverse events that could be caused by drug combinations. Signal detection algorithms analyse these databases and generate signals (i.e. evidence of potential causal relationship between a specific drug combination and one or more adverse events) that are later assessed for clinical plausibility. In a Bayesian setting, the large size of the hypothesis space when considering potential interactions between large numbers of drugs necessitates efficient numerical techniques. Hence, it would be possible to navigate the space of all the hypotheses relevant to potential DDIs hidden in a database of reports for many drugs. This research will draw upon expertise in signal processing in order to provide a Bayesian approach to identifying potential DDIs, aiming at developing novel algorithms that outperform the existing state-of-the-art and can be used in a real-world setting.
Supervisory team: Professor Simon Maskell, Professor Sir Munir Pirmohamed
ORCID: 0000-0002-4053-0220
Bio
Gemma graduated in 2013 with an undergraduate masters in Mathematics from the University of Liverpool, with her masters project focusing on the numerical solutions to problems in image processing. In 2017 Gemma returned to the University of Liverpool to study an MSc in Big Data and High Performance Computing with a project focusing on finding structurally similar crystals using crystal geometry and data mining techniques. For my PhD project Gemma will be aiming to improve the accuracy of predictions of collisions in space between satellites and orbital debris.
Electrical Engineering & Electronics and Mathematics.
Research project title: Improving conjunction analysis using a combination of high-fidelity astrodynamics models and advanced numerical Bayesian methods
Supervisory team: Prof. Simon Maskell and Dr Corina Constantinescu
Bio
After having graduated from Aerospace Engineering Master's (MEng) programme at the University of Liverpool, I decided to focus my attention on a topic where I could have the most meaningful impact. Food production is at the core of many of our economic, environmental and social challenges. As global populations continue to rise, we must increase food production by 70% to feed a projected global population of 9.77 billion by 2050, with limited arable land, whilst also attempting to reduce associated greenhouse gas emissions. This is whilst tackling the issues of water scarcity, climate change and rapid urbanisation. I therefore changed my direction towards building a sustainable future, and have undertaken a Ph.D. to research solutions for sustainable food production.
Vertical farming, sustainability, decision support systems, economic forecasting, risk profiling, montecarlo simulations, environmental impact assessment, life cycle analysis, smarter mobility, lean manufacturing.
Research project title: Developing a decision support system to enable sustainable growth of small-scale vertical farms through environmental impact and financial risk modelling.
Project description: My research focuses on modelling the sustainability of vertical farming. In particular, I aim to provide an open-source decision support system that provides economic projections for companies wishing to venture into vertical farming. It will provide the user with a risk profile based on their decisions and local parameters, as well as provide methods to mitigate risks and reduce environmental impacts. The research is a truly collaborative project, drawing on expertise from the University of Liverpool's Institute for Risk and Uncertainty, Schools of Management and Engineering, and the Institute of Integrative Biology, whilst working closely with Farm Urban, a pioneering R&D company.
Supervisory team: Prof. Scott Ferson; Dr. Ronald Dyer; Dr. Adam Mannis; Dr. Iain Young
ORCID: 0000-0001-5386-2008
Bio
Hayley Jones graduated in 2017 in Mathematics from Liverpool John Moores University, during her undergraduate degree she worked as a Data Analyst and Modeller, and before joining the Risk institute worked as a Business Analyst. Her interests are in mathematical modelling for medical research specifically using ODE’s and PDE’s.
Mathematical Modelling, Machine Learning and Clinical Medicine
Research project title: Mathematical Modelling and Image Analysis in Ocular Melanoma
Supervisors: Prof Sarah Coupland, Prof Ke Chen, Prof Azzam Taktak
BioBSc degree in Physics from the Autonomous National University of Mexico (UNAM, Mexico) in 2015. Worked as assistant researcher at the project “Multi-wired moun’s detector at the Teotihuacan pyramid” from 2012 to 2015. Awarded with a scholarship from the National Council of Science and Technology (CONACyT) and the Energy Secretary (SENER) of Mexico to study the Master's degree (MSc) in Radiometrics: Instrumentation and Modelling in 2016 at the School of Physics of the University of Liverpool (UoL). Currently, PhD student at the Institute for Risk and Uncertainty of the UoL since 2017.
Credal networks, imprecise probabilities, p-box, Bayesian networks, Risk assessment, approximate inference
Research project title: Probabilistic risk assessment of severe accidents for critical installations
Project description: In order to improve the robustness of facilities against natural threats, risk assessments must be carried out considering updates of climate conditions which can be different from those for which technological installations were designed first. This new information can be useful not only for long-term decision making but also, in the case of an emergency event. In this regard, probabilistic methods like Bayesian networks have increasingly been used to perform risk assessments proving to be a reliable and powerful tool with the flexibility to include different types of data and capturing the complexity of the system under analysis.
Credal network is a generalization of the Bayesian network approach for imprecise probabilities that is being explored as an option to represent the uncertainty due to the lack of information or conflict data. This will allow performing a robust analysis despite the imprecise conditions of the data.
Supervisory team: Dr Edoardo Patelli; Dr Marco De Angelis; Dr Francisco Alejandro Diaz de la O
ORCID: 0000-0001-5929-6414
Bio
Hindolo is currently a post doc in the Department of Engineering Science at University of Oxford. His PhD thesis is available at . Hindolo graduated with a first-class bachelor’s degree in electrical/electronic engineering from the University of Sierra Leone in 2010. Two years later, he won a commonwealth shared scholarship to pursue an M.Sc. (Eng.) degree in energy generation at the University of Liverpool, where he graduated with a distinction. His M.Sc. thesis was on the reliability and availability analysis of hydroelectric power plants, using the Bumbuna hydroelectric power plant in Sierra Leone as a case study. Symmetrically sandwiched between his undergraduate, master’s, and PhD studies are his two spells with the Sierra Leone affiliate of the French oil giant, Total. He was the country technical and administrative head of Total’s maintenance team in his second spell, and in consonance with the projects team, championed the successful implementation of several critical engineering projects. Hindolo’s key expertise and interests include, energy systems modelling, mathematical modelling and simulation of complex processes, uncertainty quantification, probabilistic risk assessment and management, reliability and maintenance optimization of complex systems, and algorithm design.
System Reliability Analysis, Maintenance Modelling, Renewable Power Generation and Integration, Resilience Modelling of Complex Systems and Infrastructure
Research project title: Efficient Reliability Modelling & Analysis of Complex Systems with Application to Nuclear Power Plant Safety
Project description: Nuclear power may be our best chance at a permanent solution to the world's energy challenges, owing to its sustainability and environmental friendliness. However, it also poses a great risk to life, property, and the economy, given the possibility of severe accidents, during its generation, transmission, and distribution. These accidents are a result of the susceptibility of the generating plants to component failure, human error, extreme environmental events, targeted attacks, and natural disasters. Given the complexity and high interconnectivity of the systems in question, a small glitch, otherwise known as an initiating event, could cascade to catastrophic consequences. It is, therefore, vital that the vulnerability of a plant to these glitches and their ensuing consequences be ascertained, to ensure that the appropriate mitigating actions are taken. A nuclear power plant by default is equipped with safety systems to deter the propagation of an initiating event. An accident ensues if the safety systems required to mitigate some initiating event are unavailable or incapacitated by the initiating event. The reliability, therefore, as well as the availability of these systems, shape the safety of the plant.
This project aims to develop a series of computationally efficient tools to address the limitations of the existing probabilistic risk assessment tools. To achieve this, the feasibility of well-established concepts like advanced Monte Carlo methods, network theory, probabilistic models, and other novel approaches will be explored. The resulting tools should influence robust decisions in the risk management of nuclear power plants. These tools will be incorporated into the open-source uncertainty quantification tool, OpenCOSSAN (see http://www.cossan.co.uk/software/open-cossan-engine.php), and therefore, readily available to industry, as well as other researchers.
Supervisory team: Edoardo Patelli; Min Lee
ORCID: 0000-0002-9316-3911
Bio
Irma graduated in Aerospace Engineering at Politecnico di Torino. My master degree concerns the design and the analysis of aerospace structures, including aeroelastic problems. My thesis project, carried at The University of Liverpool, aimed to develop the dynamic control of an aeroelastic system applied to a wing, tested in wind tunnel.
Aerospace and control engineering
Research project title: Aeroelastic Control of Suspension Bridges under Construction
Supervisors: Sebastiano Fichera, Daniel Colquitt and Stephen Jones
Bio
I received my undergraduate degree in Artificial Intelligence from University of Liverpool in 2016. I went on to study a CDT with the Risk Institute for which I received my MRes in September 2017. I first discovered Genetic Algorithms and Neuroevolution during a second year undergraduate course when a group of colleagues and I wrote a simulation that evolved fighting agents. I was completely fascinated at how a computer program could learn such sophisticated control policies. Since then I have continued to work in this area in order to really discover how powerful these algorithms are.
My research interests are mainly in the area of applying machine learning techniques such as Neuroevolution to robotic swarms. Although, I am interested in anything to do with ML, Deep Learning, Multi-Agent Systems or Decision Making under Uncertainty
Research project title: Decentralised Mapping of Indoor Hazardous Environments Using MAVs
Project description: This project aims to develop a technical, algorithmic and mathematical framework that enables a swarm of MAVs to perform decentralised mapping of hazardous environments. The main issues to tackle are the areas of coverage, navigation, vision and map fusion algorithms as these all must be addressed for high quality surveillance and mapping system. This project will look at how to apply the ideas of Genetic Algorithms and Neuroevolution to evolve control policies for the agents in the swarm.
Supervisory team: Prof Karl Tuyls; Dr Paolo Paoletti
ORCID: 0000-0002-6446-6557
Bio
I gained an MPhys in Physics with Theoretical Physics at the University of Manchester, before joining the Next Generation Nuclear Center for Doctoral Training to study for a PhD in Engineering. After one year taught course in Nuclear Science and Engineering I started my PhD research, which intersects the disciplines of Machine Learning. Nuclear Engineering, Software Engineering and Uncertainty Quantification.
Machine learning, Bad data, Imprecise probability, Safety Analysis
Research project title: Robust Probabilistic Risk/Safety Analysis for dealing with Scarce and Limited Data
Project description: The safety of nuclear systems and plants relies on the reliability and availability of different safety systems. Under realistic conditions, these systems and components are affected by uncertainties, caused by lack of sufficient knowledge and/or by natural unpredictable external events. Uncertainty analysis is an essential tool to obtain a robust representation of model predictions consistent with the state-of knowledge. Generally speaking, there are two kinds of uncertainty involved in nuclear safety evaluation, which are aleatory uncertainty and epistemic uncertainty. For instance, the calculation uncertainty of best estimate event analysis must be quantified by considering both model uncertainty and plant status uncertainty, and generally, a conservative surrogate sequence is assumed for a particular event safety evaluation. For example, in the large break loss of coolant accident analysis, the loss of off-site power and single failure are conservatively assumed in the evaluation surrogate sequence. In fact, owing to the random failure of mitigation systems or components, there are many possible event sequences which exist with different occurrence probabilities.
Probabilistic Safety Analysis has a vital role to play in safety analysis of complex nuclear plants. These techniques can be used to provide a quantitative analysis of the identified risk. However, such analyses depend on the assumptions and the failure distributions for each component. Often not enough data is available to define such distributions and some expert judgement is necessary.
The overall aim of the project is to develop a robust methodology for probabilistic safety (risk) assessment. It involves the development of an efficient general purpose computational tool and strategies able to deal with scarce, vague and imprecise information and a methodology that will provide a robust support tool for risk-informed decision making.
Supervisory team: Edoardo Patelli; Marco de Angelis
ORCID: 0000-0003-4106-2374
Bio
I am an MRes student with a background in mathematics working alongside the Institute for Financial and Actuarial Mathematics. My research focus is pension systems, with my main project focusing on the risk analysis of pension systems for survivors, with a particular interest in developing countries.
Actuarial mathematics, pension systems, mortality rates, broken-heart syndrome
Research project title: Risk Analysis of Pension Systems for Survivors (Orphans and Widows)
Project description: By building a data base of existing financial mechanisms, we aim to identify ways to improve social security systems for survivors - orphans and widows. The initial aim is to better capture the current social security landscape, so that sensible recommendations can be made for policymakers.
Supervisory team: Dr Corina Constantinescu; Professor Sandra Walklate
ORCID: 0000-0002-9018-7993
Bio
Liam Doyle is a postgraduate research student at the University of Liverpool Institute for Risk & Uncertainty. His current research is in high-throughput component testing and materials certification, with a focus on additive manufacturing and utilising imprecise probabilities.
Liam completed his undergraduate degree in mechanical engineering at the University of Birmingham, after which he spent time in industry facing challenges such as hydrogen fuel cell development and the design of aircraft tyres.
Additive manufacturing, Materials testing, Imprecise probability, Uncertainty
Research project title: High-Throughput Component Testing and Certification for Additive Manufacturing
Project description: This project aims to identify and control process parameters in additive manufacturing to obtain data for outliers enabling high-throughput testing. This would reduce the uncertainty present in additively manufactured components. Therefore enabling the uptake of additive manufacturing in highly regulated industries.
Supervisory team: P.L. Green, C. Sutcliffe
ORCID: 0000-0002-9018-7993
Bio
I am a member of the Institute for Risk and Uncertainty as a student at the Centre for Doctoral Training. My Bachelor’s degree in Mathematics is from the University of Valencia (Spain) and my Master’s degree in Decision Making under Risk and Uncertainty is from the University of Liverpool. My research aims to provide a theoretical and empirical analysis of time-varying volatility in financial markets.
My research interests include financial mathematics, time series analysis and Bayesian econometrics.
Research project title: Markov-switching GARCH models: a unifying framework
Project description: The primary goal of the research is to provide a theoretical and empirical analysis of various Markov-switching GARCH models. With this aim, we first develop a parametric family of models that nests the regime-switching version of many commonly used GARCH-type models. We then provide conditions for asymptotic covariance stationarity and therefore for finite second-order moments. Further research considers conditions for the existence of moments of any order. Our family of models is a unified framework in which various Markov-switching GARCH models can be tested. Deriving statistical properties of several models as special cases of a unifying framework allows to compare how well different models reproduce stylised facts of financial data. We will provide such comparison through empirical analysis.
Supervisory team: Brendan McCabe; Olan Henry; Carmen Boado Penas
ORCID: 0000-0003-1323-5397
Bio
Marios Filippoupolitis graduated from Patras University with a BEng and MEng in Civil Engineering. He has an MSc in Advanced Structural Analysis of Monuments and Historical Constructions from the Universities of Minho and Padova. Marios completed his MPhil in the Acoustics Research Unit at the University of Liverpool on finite element modelling of dowelled timber-joist floors that was funded by the Swiss National Science Foundation. Marios is now registered for PhD study on structure-borne sound transmission in fragmented building structures after earthquakes.
Structural dynamics, Finite Element Analysis, Statistical Energy Analysis, Experimental Modal Analysis, Structural Analysis
Research project title: Detection of trapped earthquake survivors
Project description: This research project is funded by the EPSRC and concerns an approach to search for human survivors using structure-borne sound propagation in collapsed and fragmented reinforced-concrete buildings through the development, validation and use of theoretical models.
Supervisory team: Carl Hopkins, Siu-Kui Au
ORCID: 0000-0002-5835-6582
Surrogate Modelling, Uncertainty Quantification, Stochastic Optimisation, Machine Learning
Research project title: Robust Design under Uncertainty
Project description: The goal of engineering design is to create technological systems that satisfy specific performance objectives and constraints over a period of time. However, modern engineering systems are inherently complex, and this complexity means that endogenous (geometry, material properties) and exogenous (loads) information is never complete. This lack of information can be captured by modeling uncertainties as random variables, whose distributions can in principle be obtained from expert opinion, literature or test data. The objective of performance-based design is therefore to determine the optimal design that minimizes an expected loss function which depends on both the characteristics of the design space and the model parameters that encode the characteristics of the system under study. However, simply optimizing for nominal performance usually results in a suboptimal design, as this approach fails to take into account the uncertainties that arise in modeling, manufacturing and operation. Moreover, once in operation, such designs usually suffer a sharper decline in performance due to degradation compared to other suboptimal designs. The aim of this project is to take all forms of uncertainty into account to perform robust optimization; obtaining a set of solutions with high performance and low variability in the face of uncertainty.
Supervisory team: Dr Francisco Alejandro Diaz De la O, Dr Peter Green
ORCID: 0000-0002-9018-7993
Bio
I graduated with a BSc. in physics from the University of Bristol. My dissertation looked at the future of nuclear power, including fusion, from an underlying physics perspective. Following graduation, I went on to work in engineering consultancy in the power sector. After several years I decided to pursue a career in post-graduate research. I studied a postgraduate course in Fusion Energy with the University of York.
My current research looks at risk quantification in the next step fusion device, DEMO. My goal is to quantify uncertainties in the design of DEMO, to aid in finding an optimal operating point.
My research interests include financial mathematics, time series analysis and Bayesian econometrics.
Research project title: Risk Quantification in Fusion Power Plant Design
Supervisory team: Dr E. Patelli, Dr H. Lux
ORCID: 0000-0002-6667-3254
Bio
Nestoras Chalkidis gratuated in 2014 from the Department of Mathematics of the Aristotle University of Thessaloniki, with a bachelor degree in Mathematics. In 2016 he obtained a Master of Science degree in Complex Systems and Networks, from the Department of Mathematics of the Aristotle University of Thessaloniki. His thesis with the title ”Statistical Analysis of epigenetic data for CLL patients” held of in the Centre for Research and Technology Hellas (CERTH), in Thessaloniki. In September 2017 he joined the University of Liverpool, and the Institute for Risk and Uncertainty as an MRes student.
Analysis of Financial Markets, Classification, Data Mining, Deep Learning, High – Frequency Trading, Machine learning, Random Forest, Recurrent Neural Networks, Variable Selection.
Research project title: Advances in Financial Risk Analysis, Modelling, and Management
Project description: This project is a multidisciplinary work which combines three major fields Mathematics, Machine Learning (ML) and Finance. The purpose of this project is to apply Artificial Intelligence (AI) and ML algorithms to Finance and in particular to Trading.
In fact, within the context of the research project with the title "Advances in Financial Risk Analysis, Modelling, and Management" the plan is to explore new and advanced ways of modelling financial systems. ML algorithms, AI and/or data analysis will be applied in a try to build a trading system application. A trading system is a system of rules or parameters, which determines when to buy or sell an equity in the stock market. To this end, a predictive model will be built as the foundation of a trading strategy.
For my PhD work, I will continue in the direction of using ML algorithms fed with market data. I will use richer raw financial data such as intra-day and order book data. Furthermore, I will explore a range of ML models, including Deep Learning, developing my own where appropriate. Finally, I will evaluate not just ML models, but trading strategies build on top of these models.
Supervisory team: Professor Rahul Savani, Mr. Jason Laws
ORCID: 0000-0003-1936-9986
Bio
I started my PhD in September 2018, I am working with the Digitwin project looking at ways in which uncertainty can be included in digital twins. Currently I am working on a software tool that will allow intrusive uncertainty analysis to be added to computer code automatically. I am also interested in how to express the uncertainty in situations in which you have no gold standard information and how that can be communicated. Another area that I am interested in is how computer algorithms can be designed in order to be more humane, how they should deal with risk and uncertainty and whether they can be designed to fail in a non-catastrophic way.
Statistics, Computer Science, Engineering
Research project title: Uncertainty Quantification and Propagation in Digital Twins
Project description: Quantitative management of financial risk, and the modelling of financial time series itself are two areas that are intimately related. The goal of this PhD project is to explore new, and advanced ways in managing risk, and modelling financial systems using machine learning algorithms, artificial intelligence and/or data analytics.
Supervisory team: Scott Ferson, Marco de Angelis, Ivan Au
Bio
Nikola studied Psychology and then Cognitive Science at Comenius University in Bratislava, in Slovakia*. After graduation she worked as a data scientist in Modelling and Prototyping team at Swiss Re** where she further developed her passion for analyzing data.
Epidemiology, Public Health
Research project title: Improving understanding of gastrointestinal disease using routinely collected surveillance data
Project description: Quantitative management of financial risk, and the modelling of financial time series itself are two areas that are intimately related. The goal of this PhD project is to explore new, and advanced ways in managing risk, and modelling financial systems using machine learning algorithms, artificial intelligence and/or data analytics.
Supervisory team: Professor Sarah O'Brien, Dr Helen Clough, Dr John Harris, Professor Miren Iturriza-Gomara
Bio
Noémie started her PhD along with a MRes in Decision-Making under uncertainty at the Institute for Risk and Uncertainty, University of Liverpool (2016). She currently works on maritime shipping optimisation subject to uncertain environments and has a strong interest for nonlinear dynamics forecasting. Her prior French academic journey includes a MSc in Continuum Physics and Mechanics, a MSc in Agricultural sciences and a MSc in Physics and Modelling of Complex Systems. Noémie investigated the issue of mesoscale oceanic turbulence on the occasion of two placements at the Ocean Physics Laboratory - Ifremer (France) and taught economics and business management for two years in a French farming high school. Her main interests are modelling and applied mathematics with a nature or industrial flavour.
Weather forecasting, Modelling nonlinear dynamical systems, Uncertainty modelling, Decision-making under uncertainty, Information theory
Research project title: Optimising cargo loading and ship scheduling subject to uncertain sea levels
Project description: In this project, we investigate the possibilities of cargo loading and ship scheduling optimisation by considering the forecasts of sea level anomalies in addition to the tide predictions. A shipping decision model is built, allowing to optimise the shipper's decisions given general information on the ship design, cargo and port management, as well as sea level forecasts for the ports of call. Different approaches for representing and propagating sea level uncertainty are investigated, based on the information available to the shipper, from tide predictions alone to sea level ensemble forecasts.
Supervisory team: Peter Green; Scott Ferson
ORCID: 0000-0003-1936-9986
Bio
I received a BSc in Mathematics from the University College Cork, and graduated in May 2014. Following this, I studied for a MSc in Financial Economics. I began my PhD at the University of Liverpool in September 2016 in Statistical Machine Learning.
Supervised Machine Learning, Data Visualization, Noisy Data, Uncertainty Quantification.
Research project title: A Supervised Machine Learning Approach for Structural Health Monitoring (SHM)
Project description: Broadly speaking, damage identification and SHM can be achieved either through model-driven approaches (usually finite element analysis) or data-driven approaches, via a statistical representation of the structure or system . An important type of data-driven approach in machine learning is classification, where the aim is to assign an input pattern to one member of a set of classes. Our industrial partner is interested in developing methods for SHM in off-shore wind turbines. A data-driven damage detection approach such as probabilistic classification could provide the necessary guidance in order to achieve optimum repair costs and reliable structural integrity assessment.
Supervisory team: Alex Diaz, Simon Maskell, Sui Kui Au
ORCID: 0000-0002-4075-8591
Bio
I am currently working towards my PhD in the University of Liverpool. I hold a BEng (Hons.) in Aerospace Engineering. I attained a degree in micro-processing technologies, before moving to the UK to pursue my passion in aerospace engineering.
I have experience in working with and for the industry. I enjoy both computational modelling and experimental, hands-on work.
I am interested in aerospace design, aerodynamics, computational fluid and structural modelling. The field of machine learning, inference and optimisation is my newest passion and I am keen to use it in aerospace applications. I am also a proponent of electric, hybrid-electric and non-standard propulsion for aircraft - fields that can greatly benefit from advances in computational methods.
Research project title: Numerical modelling and uncertainty quantification for biodiesel filters
Project description: The project is part of an optimal filter design framework which aims to develop a systematic approach to the design of non-woven coalescing filters. The work utilizes fluid dynamic modelling, machine learning methods and uncertainty quantification approaches.
Supervisory team: F.A. Diaz De la O; K.J. Kubiak; U Farooq
ORCID: 0000-0002-3302-686X
Bio
I started my university study in Bologna, Italy, where I obtained a bachelor degree and a master in Energy Engineering. After winning a grant, I had the chance to develop my master thesis project within an international research group at Ecole Centrale de Paris. After this experience I publish my first paperand I became passionate about research, scientific writing and computational methods in general. It did not take long before I won a second grant for a Master of Research and a PhD. The grant covered 1-year Master at Liverpool University in the field ‘Decision Making under Risk and Uncertainty’ and 3-years of PhD in the same subject. My PhD is a multidisciplinary project bridging between engineering and computer science disciplines. During the 3 years o PhD, I had the chance to attend several international conferences, publish 4 journal papers and work in direct contact with different research groups in different places in Europe (e.g. Politecnico di Milano and ETH Zurich). I developed a solid knowledge of uncertainty quantification method and probabilistic approaches. I also had the chance to work with novel and popular methods such as, for instance, Deep Reinforcement Learning and Bayesian updating methods.
Systems and Networks, Reliability, Vulnerability, Resilience, Machine Learning, Reinforcement Learning, Uncertainty Quantification, Sensitivity Analysis, Information Modelling, Simulations
Research project title: Enhance Smart Grid Reliability and Resilience
Project description: The aim of this project is to tackle problems related uncertainty and lack of data in the future grid (i.e. smart grid) resilience context. The effect of the interactions between power grid operations, its topological structure and external weather events are investigated. Power grid reliability, vulnerability and resilience concepts compared. The expected contribution is an efficient and effective computational framework which embeds a complex environmental-grid model resilience-reliability-vulnerability models. The newly developed frameworks will be capable of dealing with lack of reliable data and severe sources of uncertainty inevitably affecting the power grid system. The project is highly multidisciplinary, crossing knowledge from a variety of disciplines such as electrical engineering, reliability engineering, environmental science and computer science.
Supervisory team: Edoardo Patelli; Matteo Broggi; Sven Schewe
ORCID: 0000-0002-8117-8737
Bio
Sara holds a BSc in Anthropology from the University College London, MSc in Gender from London School of Economics and an MSc in Clinical Animal Behaviour from the University of Lincoln. Before starting her PhD, she has worked for a number of not-for-profit organisations in various research, analyst and recruitment roles. From 2010 she volunteered and later conducted research at Battersea Dogs and Cats Home, focusing on improving the welfare of shelter dogs. Sara has an experience in animal training and working with dog owners as a behaviour consultant and dog trainer.
Animal behaviour, Animal welfare, Behaviour change, Sociology of Risk, Social studies of science
Research project title: Dog Bites: Perception and Prevention
Project description: Dog bites and strikes led to 7,227 episodes of hospital admission in England and Wales between 2014-15 and a significant loss of income to businesses and individuals (Winter, 2015). Researchers commonly approach dog bites as a medical problem. As a result, what is overlooked is how an individual's social context, their work, and environment they live in shape their perceptions of risk and safety and influence their behaviour around dogs. The social context may also affect how the bite is experienced, for instance, two similar bites may have a profoundly different impact on individuals who live in different circumstances. This level of understanding is crucial for bite prevention as prevention campaigns are often based on a number of assumptions regarding: what people do around dogs before the bite, what they perceive as risk and safety, how they perceive dogs and dog bites and in what methods would prevent bites at a population level.
I use primarily qualitative methods (focus group discussions, in-depth qualitative interviews, participant-observations and document analysis) to examine the experiences and perceptions of dog bites and dog-related practices among people with different experiences of dogs and dog bites.
Supervisory team: Carri Westgarth, Francine Watkins, Rob Christley and Huadong Yang
ORCID: 0000-0001-5323-8117
Bio
I am an environmental scientist interested in predictive modelling, climate change, fluvial systems, environmental policy as well as being a passionate science communicator. My goals are to reduce uncertainties surrounding flood risk under climate change, drawing on my academic knowledge in computational modelling, monitoring and reconstruction, as well as experience in previous roles analyzing environmental policy and in environmental assessment.
Climate Change, Flooding, Environmental Hydraulics, Computational Fluid Dynamics, Policy
Research project title: Risk of river flood inundation under climate change: assessment of the relative effects of changes in plant growth and flood regime on conveyance
Project description: My project aims to enhance flood risk estimation for river systems under climate change with a special focus on aquatic vegetation, a key yet neglected component influencing river hydraulics. Greater vegetation abundances are thought to increase flood risk; climate change is expected to increase both the frequency of extreme storms and vegetation growth rates, threatening a perfect storm of higher flows exaggerated by higher vegetation abundances. Using computational fluid dynamics this projects is investigates the changing dynamic between vegetation and river flow under various climate change scenarios to estimate changes to future flood risk.
Supervisory team: James Cooper; Ming Li; Janet Hooke
Stock Markets, Algorithms, Chaos Theory, Forecasting, Causality, A.I.
Research project title: Finding Hidden Structure in Financial Networks
Project description: My target is to uncover hidden interactions between financial time series such as stocks, bonds and CDS and apply statistical arbitrage and forecasting enhancement by mining couplings that cannot be captured by industry-standard tools (i.e. correlations).
Supervisory team: Primary Supervisor: Dr. Athanasios A. Pantelous, Secondary Supervisor: Dr. Konstantin M. Zuev
ORCID: 0000-0001-5323-8117
Bio
Vladimir completed his bachelor’s degree in Artificial Intelligence and Computer Science at the University of Edinburgh in 2018. He joined the Institute for Risk and Uncertainty in September of the same year.
Research project title: Modelling Just In Time with Uncertainty
Project description: Being able to predict the limits of a production line can be beneficial in conditions with increased environmental and political uncertainty. Currently, most production flow models consider uncertainty either for the delivery of resources or the internal structure of the production flow (failure of production equipment). The aim is to present a tool that can combine sources of uncertainty and simulate the production flow under a demand function to help determine the theoretical limits of the system.
Supervisory team: Scott Ferson, Ullrika Sahlin
Emulation, Rare Event Sampling, Bayesian Analysis, Model Calibration, Robust Design, Correlation, Fuzzy Analysis, Clustering, Engineering statistics.
Research project title: Calibration of Expensive Computer Models Using Engineering Reliability Methods
Project description: My target is to uncover hidden interactions between financial time series such as stocks, bonds and CDS and apply statistical arbitrage and forecasting enhancement by mining couplings that cannot be captured by industry-standard tools (i.e. correlations).
Supervisory team: Francisco Alejandro Diaz De La O; Michael Beer
ORCID: 0000-0002-0118-7108
Bio
Zuo Zhu is a PhD student in Institute for risk and uncertainty. He obtained BEng and MEng from Dalian University of Technology, all in civil engineering. His current research focuses on modal identification, Bayesian methods and structural dynamics.
Modal identification, Bayesian methods and Structural dynamics.
Research project title: Structural modal identification with application in civil engineering
Supervisory team: Prof. Siu-kui Au; Dr. Peter Green
ORCID: 0000-0002-5025-6640