Method for discovering frequent sets in databases.; We present a method to discover patterns of frequent appearance in a database. From these patterns you can be deduced association rules in the appearance of items in the database. The basic idea is to find the item sets with frequency of occurrence above a predetermined threshold, using a generator space search the contents of the database itself and using the information gleaned from the search itself to reduce the maximum exploration of possibilities that will not be part of the solution. This method may perform scanning of complex databases with high level of detail locating associative rules with low support (low frequency), but of high reliability and significance in application environments as bases gene or other biological data.
展开▼