Research Group on Forecasting, Data Mining and Market Modelling

Forecasting and market modelling are important activities in organisations. Few organisations can survive for long if they fail to make accurate forecasts, nor understand the key market drivers of their business. 

Forecasts are made at the operational level in organisations to support the day-to-day activities such as scheduling staff, production and logistics. With a longer time frame, marketing requires the development of plans and forecasts that depend on such drivers as price, promotions etc. Within the marketing area, new product forecasting is particularly prone to error with often disastrous financial consequences. At the level of the individual consumer, organisations need to understand how consumers value their products and services and how they respond to marketing initiatives. Research in all these area is valuable to organisations and consultancies, which seek to employ those with expertise. There is also a high demand for research academics in both these two fields.

Regular PhD opportunities

We always looking for PhD students interested in the areas of time series analysis, forecasting, data mining and market modelling in joining our group. A full list of topics may be found on the departmental pages on PhD positions in forecasting.

Potential PhD topics in forecasting (not limited to these!)

  • Statistical methods for forecasting to support supply chain management
  • Forecasting with computational intelligence
  • Forecasting with artificial neural network
  • Forecasting with support vector regression
  • The evaluation of econometric principles for improved forecasting
  • Forecasting in telecoms and the adoption of new services
  • Evaluating and interpreting macroeconomic forecasts
  • Evaluating the relative forecasting properties of non-linear time series models
  • Forecast evaluation from the perspective of asymmetric loss functions

Potential PhD topics in data mining (theory and applications)

  • Dynamic data mining, adaptive data mining (for data with distribution drift)
  • Data mining applications in business and supply chain management
  • Methods of computational intelligence for data mining, in particular neural networks and support vector machines
  • Cost sensitive learning, active learning, semi-supervised learning
  • Data mining on time series data (time series clustering and time series classification)

Potential PhD topics in data mining, market modelling and predictive analytics

  • The evaluation of market share models
  • Brand value and brand choice
  • Short- and long-run marketing effectiveness
  • Optimal marketing resource allocation
  • Customer lifetime value
  • Marketing-finance interface

The academic staff members associated with this research area are Professor Robert FildesDr Sven F CroneDr Nikolaos Kourentzes and Dr Nicos Pavlidis. Students working in forecasting can also draw on the expertise of additional staff associated with the Centre for Marketing Analytics and Forecasting. There are regular informal discussion meetings in both marketing and forecasting where students can test out their ideas or discuss the latest research papers.

For informal enquiries please contact