A modified approach to mine frequent patterns from uncertain data

There are so many existing algorithms proposed that mines frequent patterns from certain or precise data. But now a days demand of uncertain data mining is increasing. There are many situations in which data are uncertain. There are mainly two approaches for mining uncertain data like level-wise approach and pattern-growth approach. Level-wise approach uses generate and test framework so it requires multiple scan such as U-Apriori algorithm. Pattern-growth approach uses tree like structure so it requires large memory and more computational time such as UF-growth algorithm, UFP- growth algorithm, and PUF-growth algorithm. From the all existing algorithm PUF-tree algorithm is more compact so it requires less memory then other existing pattern-growth approach based algorithms. In procedure of PUF-tree algorithm first it generates the tree and in second step it applies PUF-mine algorithm to the generated tree to mine frequent pattern from it. But the whole process is very complex and requires longer time to execute. So we can propose the algorithm that reduces the time complexity of PUF-tree algorithm by using linear list data structure instead of tree structure. Motivation behind using linear list is to minimize time complexity and make the algorithm simpler and easily understandable.