The thesis strongle analysed the Fp-growth algorithm of Apriori process,then it provided a database anomaly detection system based on Fp-growth algorithm and at last,the thesis given the results of system test data.

This method is applied to the measured steady data of a large -scale liquid propellant rocket engine and tested with FP-Growth arithmetic.

An improved FP-growth algorithm based on aggregative chains is proposed.

In order to improve the speed of mining user frequent behavior pattern and FP-tree space utilization,thereby significantly improving the efficiency of security audit data analysis,based on the FP-growth algorithm this paper proposes an improved correlation algorithm suitable for the analysis of the security audit data.

In order to adapt to the real-time nature of the intrusion of testing requirements,and enhance the efficiency of algorithms,presented to the FP-Growth algorithms that improves FP-tree table structures and the introduction of key attributes to intrusion detection system.

FP-growth is the algorithm mining frequent itemsets without generating candidate itemsets,and it is important to analyze and implement it.

The algorithm builds frequent patterns directly as FP-growth algorithm does.

It is a transformation of FP-Growth.