My main research focus is on business forecasting and data mining with artificial neural networks. I am actively engaged in both international forecasting associations; as the official tutor for the 2004 & 2005 tutorial on forecasting with artificial neural networks for the IBF (Boston, & Chicago) and Chair of the special track on neural networks for forecasting at the IIF conference. In addition, I am webmaster for the neural forecasting portal (at www.neural-forecasting.com).
Reserach interrest are centered around my doctoral research on Forecasting with artificial neural networks. More specifically, the use of artificial neural networks for simultaneous forecasting and safety stock calculation in inventory managent using asymmetric objetcive functions motivated from marginal costs of the managerial decision process. General research directions encompass Management Science, i.e. Operational Research and Information Systems in the domains of Demand Planning within Inventory Management for Supply Chain Management & Operations Management.
Current research extends in two mayor directions: Forecasting and Data Mining.
- Forecasting --> Methods, applications, processes and information systems for Demand Planning in Advanced Planning Systems (APS) and Enterprise Resource Planning (ERP) Systems
- Neural Networks --> Methods & applications for fully automatic Forecasting, e.g of seasonal, intermittent and generally instationary time series (outlier and level shift infused)
- Neural Networks --> Objective Functions for Neural Networks & Computational Intelligence in forecasting and Data Mining
- Neural Networks --> Methods & applications for Data Mining, Predictive Analytics and Business Intelligence
I am also extending my reserach in Management Science, Information Systems & Operations Reserach from forecasting towards more general issues of demand planning in supply chain management and other areas of forecasting.
For more information on ongoing, applied reserach please visit www.neural-forecasting.com
since 2005 Information Portal on Forecasting with Neural Networks
Creator, developer and webmaster of www.neural-forecasting.com, an interdisciplinary online portal on neural forecasting. The portal attracts 4,000 unique visitors / 10,000 page loads per month, and has build a specialist community of over 800 academics, practitioners and students interested in forecasting with neural networks.
since 2003 Software Simulator for forecasting with Neural Networks
Development of an academic software package for large scale empirical experiments with neural networks for forecasting. The software is developed using a rapid prototyping approach for over 3 man-years. Since 2005 it is supported through an externally funded, full-time Research Assistant for software development.
RESEARCH Grants & Funding
2005/2006 SAS & IIF Forecasting Research Grant
Joint grant awarded by SAS and the International Institute of Forecasters for a large scale empirical forecasting competition on neural network forecasting in 2006/07. Results are to be presented at the 2007 International Symposium on Forecasting in New York, and published in the International Journal for Forecasting, IJF
(jointly with K. Nikolopoulos, University of Manchester, UK).
2004/2005 Lancaster University, recipient of priming grants
Funding for high performance hardware and software for neural network experiments in forecasting and data mining.
2003 Joint Research Fellow Grant
Universities of Stellenbosch & University of Hamburg joint funding for academic research fellowship to visit USB.