Association Rules Algorithm In The Research And Application In The Classification Of Coastline

With the development of remote sensing technology and explosive growth of the ocean data, using computer technology, especially the data mining technology has become the primary means to discover useful information from massive data. As one of the important topic of marine protection, coastline classification has attracted a growing number of researchers. How to find out the coastline classification rules from the large-scale ocean data is a critical problem need to be solved.In order to improve the association rules algorithm to handle massive data efficiently, researchers proposed array-based association algorithm and the partition-based association algorithm. Array-based association algorithm stores the data in database to a two-dimensional array in memory, will scans the two-dimensional array instead of the database scanning. As it needs to scan the database only once, so efficiency is improved. However when facing of massive data association mining, due to memory limitations, it can not set entire data of the database into the two-dimensional array, making the mining process can not be carried out, the partition algorithm is a good solution to this problem. This algorithm scans the database three times, and will get all of the association rules of the entire data finally. But when facing huge amounts of data, scans the database three times is a serious problem that will impact the efficiency of the algorithm.Firstly, proposed partition Apriori algorithm based on one time scanning. Putting random samples into each partition, then the associated rules and the support degree of association rules in each partition were obtained through the array-based Apriori algorithm. Based on the minimum partition support degree, the approximate associated rules of the dataset were obtained. The experimental results verify the effectiveness of the proposed algorithm when dealing with massive data.Secondly, partition Apriori algorithm based on one time scanning was used for coastline type classification. The association rule is found for four kinds of coastline type recognition using samples of coastline. This association rule algorithm can be used to distinguish the four kinds of coastline in massive data.