An Improved Algorithm for High Speed Train´s Maintenance Data Mining Based on MapReduce

The daily maintenance of the high speed Electric Multiple Units (EMU) trains generates a large amount of data which can be utilized for EMU´s fault diagnosis. The existing parallel frequent pattern growth algorithm for data mining has some weaknesses in this application. In this paper, an improved algorithm is proposed by using the local frequent pattern tree (FP-Tree) instead of the global FP-Tree. This algorithm adopts parallel processing in every data processing steps. The production rules of the local FP-Tree are optimized, and the searching strategy of the frequent patterns are also improved. This algorithm proves fast, highly efficient and accurate in the experiments in the process of EMU´s fault diagnosis.