Research on Self-Adaptive Stream Data Mining

Abstract

This paper presents a self-adaptive stream data mining algorithm. The modification of the currently existing outlier stream data mining algorithm can make the stream data mining algorithm have adaptability, thus adjusting to the changes in the nature of the stream data caused by the environmental changes. The research starts from two aspects to enhance the adaptability of the stream data mining algorithm: In terms of the adaptability of system resources, the algorithm implemented in this paper can dynamically adjust the system resources used by the algorithm according to the load of the host, thus giving play to the initiative of algorithm to a maximum extent; in terms of the adaptability of stream data content, this paper researches the characteristics of the stream data, and modifies the existing data mining algorithms, so the stream data mining algorithm can make a dynamic adjustment according to the characteristics of the stream data, thus enhancing the efficiency of mining. Finally, this paper carries out comparison with the FODFP-Stream and LOF algorithm, and compares the time consumption with space consumption through the experimental data.