Various methods are used to extrapolate time series and forecast the future values of an observed variable. Most of them are based on some serious assumptions and require the user to be familiar with them. The approach used here is completely different. It falls more into a pattern recognition category, or a data mining approach. It searches for the patterns in the time series through a cluster of unobserved and artificially created vectors. Once patterns are matched, the distances between them are measured and they are used for predicting the future of the time series.

To illustrate the approach, a simple Excel programme, containing macros, was created. At present the first 40 observations of the famous Lorenz attractor are entered in the cells B5:B44. To predict the future values, just click on the button with the time series picture on it, respond to the two dialogue boxes which will emerge and see what happens next. If you want to try your own data, copy your own data into column B over the existing data set. If your data set is longer, there is no problem. If it is shorter, then delete the old values from this column. Click on the button to produce forecasts. To start the spreadsheet, click here. Don't forget, you must enable macros in your Excel.