Design and realization of outlier mining algorithm in data-intensive computing environment

Design and realization of outlier mining algorithm in data-intensive computing environment

Chenyali Changrombo Li Yahong Zhang Shusen Liu Xi

Based on the MapReduce model, a new algorithm for outlier mining based on LOF method and grid technology is proposed Mr_lof. The map phase uses the grid for data reduction to send the representative point information to the master node, and the reduced phase uses the density based outlier mining algorithm to filter out the dense regions with the help of grid value E. The algorithm only needs to compute the LOF value of the sparse area object and reduce the time complexity of the algorithm. Experimental results show that this method can effectively excavate outliers in data-intensive computing environment.

Design and realization of outlier mining algorithm in data-intensive computing environment