MLP Networks Applied to the Problem of Prediction of Runtime of SAP BW Queries

The SAP BW is a BI tool used daily by about 8000 employees of a big oil company in Brazil, running monthly about 150,000 queries to assist in the analysis inherent in their professional activities. A query is created to meet a need for specific business analysis and its response time is directly affected by the use of BW server by other users. The main problem today is that there is no way to estimate the execution time in advance, for the user to decide the best time to execute the query he needs for his work. This article proposes a solution to this problem by developing a prediction classification model for the performance of BW queries at certain times of the day using Multilayer Perceptron (MLP) neural network and the Weka tool [WEKA, 2011].