James Taylor is Professor of Decision Science at the Saïd Business School. His research is in the area of time series forecasting, and his work has appeared in a variety of academic journals. He currently teaches analytics courses for the Oxford MBA and Executive MBA Programmes.
James has approximately forty published and accepted research papers to his name (full list here). The journals in which he has had papers published include the International Journal of Forecasting, Journal of the American Statistical Association, Management Science and Monthly Weather Review. James is a former Associate Editor of the International Journal of Forecasting and Management Science. He has carried out a number of consultancy and research projects in the energy sector.
He has an undergraduate degree in Maths from the University of Cambridge, an MSc in Operational Research from the University of Lancaster and a PhD in Time Series Forecasting from London Business School. He taught at London Business School for three years before joining Saïd Business School in 1999.
Key words for his research interests include:
James’ research is in the area of time series forecasting, with a particular focus on two methodological areas: (1) the estimation of forecast uncertainty and (2) exponential smoothing methods. His work is mainly applied to the following areas: call centres, energy, finance, and inventory control.
Examples of his recent research include:
Arora, S., Taylor, J.W. 2013. Short-term Forecasting of Anomalous Load Using Rule-based Triple Seasonal Methods. IEEE Transactions on Power Systems, 28, 3235-3242.
Numerous methods have been proposed for forecasting electricity load for normal days. Far less attention has been paid to the modelling of anomalous load, occurring on special days, such as public holidays. This paper is concerned with forecasting for lead times up to a day-head on such special days. Anomalous load conditions pose considerable modelling challenges due to their infrequent occurrence and significant deviation from normal load. To overcome these limitations, we adopt a rule-based approach, which allows incorporation of prior expert knowledge of load profiles into exponential smoothing and autoregressive moving average models.
Jeon, J., Taylor, J.W. 2012. Using Conditional Kernel Density Estimation for Wind Power Forecasting. Journal of the American Statistical Association, 107, 66-79.
Of the various renewable energy resources, wind power is widely recognized as one of the most promising. The management of wind farms and electricity systems can benefit greatly from the availability of estimates of the probability distribution of wind power generation. In this paper, we focus on prediction up to three days ahead. We develop an approach to producing density forecasts for the wind power generated at individual wind farms. The approach is based on conditional kernel density estimation.
Taylor, J.W. 2012. Density Forecasting of Intraday Call Centre Arrivals Using Models Based on Exponential Smoothing. Management Science, 58, 534-549
A key input to the call centre staffing process is a forecast for the number of calls arriving. Forecasts of the probability density function of the arrival rate are needed for analytical call centre staffing models. To produce such density forecasts, we develop a Poisson count model, with gamma distributed arrival rate, which has the features of an exponential smoothing model adapted for modelling the intraday and intraweek cycles in intraday call centre arrivals data.
James is a former Associate Editor of the International Journal of Forecasting, and Management Science.
He has carried out a number of consultancy and research projects in the energy sector. These have related to various aspects of short-term electricity demand forecasting. James is also an External Examiner for the MBA Programmes at London Business School, and for the Management Science undergraduate courses at Warwick Business School.
James currently teaches the Decision and Data Analytics core course to MBA and Executive MBA students. He also teaches a similar course for the Master of Public Policy at the University’s Blavatnik School of Government. Periodically, he teaches the Statistical Research Methods course to the Business School’s DPhil students, and has previously taught an elective on Time Series Forecasting to students on MBA and Finance Masters courses.
The analytics courses that James teaches have strong links to other business school subject areas, especially finance, marketing and operations management. An emphasis of his teaching is practical application and user-level understanding. For example, Excel is used widely to illustrate how to apply the methods introduced. Beyond his work at Saïd Business School, James has taught short courses in analytics and forecasting at a number of organisations, including BSkyB, General Electric, National Grid Company, Oxford University Press, and Tesco.
Saïd Business School
University of Oxford
Park End Street