The Research Of Association Rules And The Application In The Work Of Renovating Railway Tunnel

Association rule mining is one of the most important models of data mining, which can be applied to many fields and act very well. This paper will study the algorithm of association rule mining, give a more efficient association rules algorithm, and apply it to the tunnel disease data.First of all, this paper in-depth studies the algorithms about association rule mining and the problems in the work of railway tunnel renovation. Among the association rule mining algorithms, some algorithms with multiple minimum supports can mine the rules that obey the laws of nature better, but they are more complex. Generally, the more data to mine the lower mining speeds, for they need to generate more candidate item sets or more complex tree. In terms of renovating tunnel, many useful theories are proposed, but the cause of runnel disease is so complex that most of theories can only be use to some certain diseases or engineering. So the results of renovating tunnel are not satisfactory or acceptable.And then, this paper proposes an association rule algorithm DPCFP-growth to resolve the problem of tunnel renovation. Firstly, the algorithm divides the different data into different subdata bases according to the ability of the computer. Each subdata base includes a branch of CFP-tree in the CFP-growth algorithm. Secondly, the algorithm mines each subdata base and obtains local frequent patterns. Thirdly, the global candidate frequent patterns are generated by unioning all the frequent patterns. At last, check individual frequent pattern to achieve global frequent patterns. So the algorithm reduces the I/O overhead significantly, which makes it possible to mine large database.After that, this paper analyses the characteristic feature of tunnel disease data, and finds that the algorithm DPCFP-growth can discover plenty of insteresting knowledge which can be used to guide tunnel renovating work.At last but not the least, the algorithm DPCFP-growth is applied to tunnel intelligent decision support system (IDSS) to offer the schemes how to forest and renovate diseases for each tunnel.