Frequent pattern mining-based sales forecasting. (English) Zbl 1353.91037
Summary: We decompose a time series into its trend, seasonal and irregular components and use multiple forecasting models for each component. This leads to \(\sim 100,000\) forecasters (triplets comprising T, S and I models). We design a set of heuristics based on frequent pattern mining to intelligently discover a set of consistently “good” and “bad” forecasters during the training phase. We employ and compare two strategies for forecasting – using the set of good forecasters in the testing phase versus filtering out the bad forecasters and using the resulting set of forecasters. We experiment with different training periods and support/confidence levels. We further zoom in on the set of good forecasters to identify frequently occurring good T, S and I models and use these to forecast each component series independently. Experimental results with over 30 sales series indicate that our heuristics greatly out-perform those such as the simple mean or ‘best forecaster from the past’.
MSC:
| 91B84 | Economic time series analysis |
| 62M20 | Inference from stochastic processes and prediction |
| 62H30 | Classification and discrimination; cluster analysis (statistical aspects) |
| 62M10 | Time series, auto-correlation, regression, etc. in statistics (GARCH) |
| 62P20 | Applications of statistics to economics |
Keywords:
sales forecasting; times series; decomposition; combining; data mining; frequent pattern mining; MAPEReferences:
| [1] | Agrawal, R., Imielinski, T., Swami, A.: Mining association rules between sets of items in very large databases. Proceedings of the ACM SIGMOD Conference on Management of data 1993, 207-216 (1993) · doi:10.1145/170036.170072 |
| [2] | Scott, AJ; Scott, AJ (ed.), Combining Forecasts (2001), Norwell, MA · Zbl 1264.62087 |
| [3] | Azoff ME (1994), Neural Network Time Series Forecasting of Financial Markets, John Wiley and Sons |
| [4] | Bates, J.M., Granger, C.W.J.: Combination of forecasts. Operations Research Quarterly 20, 451-468 (1969) |
| [5] | Brockwell, P.J., Davis, R.A.: Time series: Theory and Methods, 2nd Edn., Springer International Edition (1991) · Zbl 0709.62080 |
| [6] | Clemens, R.T.: Combining forecasts: A review and annotated bibliography. International Journal of Forecasting 5, 559-583 (1989) · doi:10.1016/0169-2070(89)90012-5 |
| [7] | Granger, C., Ramanathan, R.: Improved methods of combining forecasts. Journal of Forecasting 3, 197-204 (1984) · doi:10.1002/for.3980030207 |
| [8] | Hand, D.J.: Mining the past to determine the future: Problems and possibilities. International Journal of Forecasting 25(3), 441-451 (2009) · doi:10.1016/j.ijforecast.2008.09.004 |
| [9] | Ting, J., Fu, T.-c., Chung, F.-l.: Mining of Stock Data: Intra- and Inter-Stock Pattern Associative Classification, DMIN’2006. 30-36 (2006) |
| [10] | Kumar, M.: Using Clustering to Improve Sales Forecasts in Retail Merchandizing. Annals OR 2010, 33-46 (2010) · Zbl 1185.62116 · doi:10.1007/s10479-008-0417-z |
| [11] | Mahfoud, S., Mani, G.: Financial forecasting using genetic algorithms. Applied Artificial Intelligence 10(6), 543-565 (1997) · doi:10.1080/088395196118425 |
| [12] | Makridakis, S., Wheelwright, S., Hyndman, R.: Forecasting methods and Applications, 3rd edn. Wiley, New York (1998) |
| [13] | Zaki, M.J., Hsiao, C.J.: CHARM: An efficient algorithm for closed association rule mining 2nd SIAM International Conf. on Data Mining, 457-473 (1991) |
| [14] | Marahaj, E.A., Inder, B.,: Forecasting Time Series from Clusters, Econometrics and Business Statistics Working Papers 9/99, Monash University, Department of Econometrics and Business Statistics. (1999) |
| [15] | Menezes, B., Seth, A., Singh, R.: Can a million experts improve your sales’ forecasts? European Symposium on Time Series Prediction, Helsinki, Finland. Feb. 7-9, 2007 (2007) |
| [16] | Baker, M.J.: Sales Forecasting, The IEBM Encyclopaedia of Marketing. International Thompson Business Press 1999, 278-290 (1999) |
| [17] | Yaik, O.B., Yong, C.H., Haron, F.: Time series prediction using adaptive association rules, Proceedings of the First International Conference on Distributed Frameworks for Multimedia Applications (DFMA’05) (2005) |
| [18] | Smith, J., Wallis, K.F.: A Simple Explanation of the Forecast Combination Puzzle. Oxford Bulletin of Economics and Statistics 71, 331-355 (2009). doi:10.1111/j.1468-0084.2008.00541.x · doi:10.1111/j.1468-0084.2008.00541.x |
| [19] | Timmermann, A.: Forecast Combinations. In G. Elliott, C.W.J. Granger, A. Timmermann (eds.), Handbook of Economic Forecasting, Elsevier Press (2006) |
| [20] | Vijayalakshmi, M., Menezes, B.: Using Data Mining to Identify a Consistent Set of Experts for Times Series Sales Forecasting, AIA 2010, (2010) http://www.actapress.com/Abstract.aspx?paperId=37731 |
| [21] | VenuGopal: Forecasting using consistent experts. Master’s Thesis, Kanwal Rekhi School of Information Technology, Bombay (2007) |
| [22] | Viger Philippe Fournier, SPMF- A Sequential Patter Mining Framework, http://www.philippe-fournier-viger.com/spmf/ |
| [23] | Zou, H., Yang, Y.: Combining time series models for forecasting. International Journal of Forecasting 20(1), 69-84 (2004) · doi:10.1016/S0169-2070(03)00004-9 |
| [24] | Economagic.com: Economic time series page. http://www.economagic.com/. |
| [25] | Hyndman Rob. Time series data library. http://www-personal.buseco.monash.edu.au/_hyndman/TSDL/. · Zbl 0771.62070 |
| [26] | Essential Guidelines on Seasonal Adjustment -http://epp.eurostat.ec.europa.eu/cache/ITY_OFFPUB/KS-RA-09-006/EN/KS-RA-09006-EN.PDF |
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.
