| Abstract [eng] |
This work is related to the data mining process, Business Intelligence technology and forecasting algorithms. This paper analyzes the forecasting algorithms, their evaluation methods and data mining technologies. There are a number of forecasting algorithms. Often there are questions how to choose the algorithm for prediction and how to prepare for work with a particular prediction task. This paper describes an experimental model which allows to assess and accelerate the re-analysis of different algorithms for prediction tasks. The model is based on the DMX, SQL templates, and assessment methods for analysis of algorithms. According to this model, the experimental system for the analysis of forecasting methods is created through the Business Intelligence technology. Using this system, this work shows how to select and prepare a forecasting algorithm to work with a specific patients classification task. |
