STOR-i Workshop on Prediction and Optimisation, Lancaster University, 16th-17th June 2022
Analytics solutions typically involve two main tasks among others:
They are now widely popularised as predictive and prescriptive analytics, but they are usually conducted separately.
Recently, researchers have looked at the possibility of more integrated approaches for prediction and optimisation.
An exchange of ideas on the intersection of prediction and optimisation
Raise awareness of recent developments in the area
Stimulate discussion and collaboration between academics and practitioners
Any specialist in operations research and statistics is welcome
Practitioners with interests in areas of prediction and optimisation
Get in touch with us to get a registration bursary
The workshop will be held at Lancaster University, UK:
Postgraduate Statistics Centre
We will send you details how to get there, when you register for the workshop.
Information about the talks at STOR-i POP on Thursday
Keynote: Hidden gems in supply chain demand forecasting
Speaker: Juan Ramón Trapero Arenas
University: Castilla–La Mancha, Spain
Trust and fairness: contracting on a shared-based platform
Speaker: Ying Yin
University: Leiden, Netherlands
Promotion plan optimization
Speaker: Vittorio Maniezzo
University: Bologna, Italy
The data-driven newsvendor: extending a method of Ban & Rudin
Speaker: Congzheng Liu
University: Lancaster, UK
How to conclude a suspended sports league
Speaker: Ali Hassanzadeh
University: Manchester, UK
Using big data analytics to forecast governance performance
Speaker: Georgios Karamatzanis
University: Durham, UK
Tea & Coffee break
Optimizing with forecasts for a medical device manufacturer
Speaker: Cigdem Gurgur
University: Purdue Fort Wayne, US
Keynote: Robust optimization approaches for sustainable energy infrastructure
Speaker: Aurelie Thiele
University: Southern Methodist, US
Information about the talks at STOR-i POP on Thursday
Keynote: Prediction in algorithmic behaviour
Speaker: Patrick de Causmaecker
University: Leuven, Belgium
Electric vehicle charging infrastructure in urban areas
Speaker: Simon Weekx
Tea & Coffee Break
A method for the nonstationary inventory problem under correlated demand
Speaker: Mengyuan Xiang
University: Xi’an-Jiaotong, China
An ensemble approach for robust predict-and-optimise
Speaker: Egon Persak
Mitigating reporting bias in predictor-optimiser interactions
Speaker: Trevor Sidery
Company: Tesco PLC
Data-driven optimisation for maritime fleet management
Speaker: Çağatay Iris
Challenges in discount strategy optimisation across a large retailer
Speaker: Aleksandar Kolev and others
Company: Tesco PLC
Advanced planning models and solutions for large-scale freight transportation
Speaker: Fran Setiawan
Closing and farewell
16th and 17th of June 2022
Abstract: Demand forecasting is a strategic process within supply chain management. Although, usually, the main discussion of the forecasting process deals about which forecasting model (or machine learning/IA counterpart) is more accurate in terms of point forecasts, in this talk, I will be focused on three forecasting problems of great importance in supply chain management that, in my view, they have not drawn enough attention. We will start with the problem of how information sharing of end demand between companies of a supply chain can enhance the demand forecasting accuracy of upwards supply chain members. In this case, we will analyse the forecasting approaches employed to incorporate information from different echelons in a supply chain to improve the demand forecasting accuracy. The second hidden gem will be the utilization of parametric and non-parametric forecasting approaches, typically used in finance applications, to optimize the safety stock size. Finally, the third hidden gem will explore the impact of the lost-sales inventory assumption on the demand forecasting exercise. In this case, we will connect this problem with the traditional statistical literature of censored estimation, and we will propose an engineering tool as the Tobit Kalman Filter as a general solution.
Bio: Juan has an industrial engineering degree (2003), MBA (2004), and Ph. D. with an European doctorate mention (2008) from University of Castilla-La Mancha. He has been a visiting researcher at Paris Descartes University in France (June 2007-September 2007) and a posdoctoral researcher at Lancaster University Management School in UK with a Marie-Curie Intra-European Fellowship (April 2008-August 2010). Currently, he is working as a Full Professor at the business administration department at Facultad de Ciencias y Tecnologías Químicas in University of Castilla-La Mancha. He is co-director of the research group Predilab (Predictive Analytics Laboratory). His research interests are focused on forecasting problems related to supply chain management and energy systems. Within supply chain, he has experience in supply chain demand forecasting; impact of collaborative mechanisms on the bullwhip effect; assessment of judgmental forecasting; calculation of the safety stock; and development of supply chain simulation software. Regarding energy system he is interested in load electricity forecast at short and mid-term horizons; price electricity forecast at short-ter
Abstract: In this talk we present robust optimization approaches for sustainable energy infrastructure. First, we develop robust portfolio models of clean and renewable energy production over time to meet local and state governments’ short-term and long-term clean energy goals, incorporating concepts from project finance to finance the building of critical new infrastructure in the presence of demand uncertainty and price uncertainty. We model and mitigate a wide range of risks both pre-completion and post-completion to create a comprehensive strategy to achieve renewable energy goals. Second, we also design tractable, robust approaches for vehicle fleets to achieve portfolio-wide emission goals in line with the new Greenhouse Emissions Standards put forward by the U.S.’s Environmental Protection Agency in December 2021. We use customer choice models to create portfolios of new vehicles that best allow car manufacturers to fulfill the preferences of various customer segments while addressing the climate crisis. Further, we investigate various financial incentives. Our models are relevant for state and local policymakers as well as individual customers.
Bio: Aurélie C. Thiele is an Associate Professor in the department of Operations Research and Engineering Management at Southern Methodist University. Prior to joining SMU, she was an Assistant Professor and (tenured) Associate Professor of Industrial and Systems Engineering at Lehigh University in Bethlehem, PA, where she also served as the co-director of the Master of Science in Analytical Finance. She has received first prize in the George Nicholson paper competition, an IBM Faculty Award and multiple National Science Foundation grants for her research in decision-making under high uncertainty. Two of her research papers have been cited more than 500 times. Further, Dr. Thiele was twice a Visiting Associate Professor at the University Paris-Dauphine in Paris, France. At SMU she has also served as President of the Faculty Senate (2020-2021). Dr. Thiele received her “diplome d'ingénieur,” summa cum laude, from the Ecole Nationale Supérieure des Mines de Paris (Mines ParisTech) in France and her M.S. and Ph.D. in Electrical Engineering and Computer Science from M.I.T.
Abstract: We overview aims, techniques methodologies and recent developments of techniques from data science and machine learning that allow to predict and improve instance specific algorithmic behaviour. We distinguish on-line techniques, that is, data science techniques integrated into advanced algorithms, off-line techniques which can be used to improve, select or construct algorithms as well as techniques that consider the problem as living in a space of which the dimensions are set by specific properties of its instances. We discuss some examples of recent results obtained for specific problems. We give an example where a recent theoretical result for a combinatorial optimization problem provides new insights that can be used to locate a region of hard problem instances in the instance space, and hence provides insight in what impacts the expected long term behaviour of an algorithm package in a real world context.
Bio: Patrick De Causmaecker obtained his PhD at the Institute for Theoretical Physics of the University of Leuven in 1983 under the supervision of Raymond Gastmans. From 1994 on he has successfully conducted research in heuristic optimization and constraint solving, especially in scheduling and rostering. He is especially interested in structures and quantitative methods to gauge and improve the performance of optimization algorithms. From October 2005 onwards he was a full professor with the department of Computer Science at KU Leuven, Campus Kortrijk. Together with his colleague Greet Vanden Berghe, he created the CODeS Group within the department of Computer Science. In 2021, he became a Professor Emeritus. He chairs the European Working Group ‘Data Science meets Optimization’ which supports activities at EURO and several workshops such as the one at IJCAI in 2022. He is on the steering committee of the PATAT series of conferences, and co-organizes PATAT 2022, August 30-September 2, in Bruges, Belgium.
You will need to book your own accommodation. There are two options for the accommodation at the Lancaster University:
Both are well situated, not far from where the workshop will take place. There are other options in Lancaster city centre, including The Sun and the Royal Kings Arms Hotels. The town centre is approximately 3 miles (5 km) from the university campus. Buses from the town centre to campus run frequently.
Dr Ahmed Kheiri is a Senior Lecturer (equivalent to Associate Professor) in Operations Research at Lancaster University Management School, Department of Management Science and actively involved in two of its research groups: Optimisation Research Group and Health Systems Research Group. In addition, he is a member of the multidisciplinary Data Science Institute, Centre for Transport and Logistics, Lancaster Intelligent, Robotic and Autonomous Systems Centre, Centre for Health Futures and STOR-i, an EPSRC-funded Centre for Doctoral Training in Statistics and Operational Research with substantial industrial engagement.
Dr Killick is Associate Professor in Statistics at Lancaster University, UK. Her primary research interests lie in the development of novel methodology for the analysis of univariate and multivariate nonstationary time series models; changepoints and locally stationary models. This covers many topics including developing models, model selection, efficient estimation, diagnostics, clustering and prediction. Rebecca is highly motivated by real world problems and has worked with data in a range of fields including Bioinformatics, Energy, Engineering, Environment, Finance, Health, Linguistics and Official Statistics.
Adam's research is in optimisation, i.e., finding the best solution to problems that have a huge (possibly infinite) number of solutions. Optimisation is an inter-disciplinary subject, lying at the interface between Operational Research, Computer Science, Applied Mathematics and Engineering. Adam concentrates mainly on methods for solving optimisation problems to proven optimality, rather than heuristic methods. He has a particular interest in combinatorial optimisation problems, i.e., problems in which variables are restricted to take integer (whole-number) values.
Anna-Lena Sachs is a Lecturer in Predictive Analytics at Lancaster University and a member of CMAF. Her research focuses on inventory management, the intersection to forecasting and behavioural operations management. With her research, she has informed the practices at companies in different industries, in particular, in retailing, transportation and healthcare. She has developed several quantitative models to support decision makers and make better inventory decisions. She conducts laboratory and field experiments to analyse how human decision makers actually behave and what type of decision support is useful in practice.
Ivan is a Lecturer of Marketing Analytics at Lancaster University, UK and a director of Marketing of CMAF. He has PhD in Management Science from Lancaster University. His main area of interest is developing statistical methods for forecasting. He is a creator and maintainer of several forecasting- and analytics-related R packages.