Yazdan Takbiri - Department of Computer Engineering, Karaj Branch, Islamic Azad University, Karaj, Iran;
Amineh Amini - Department of Computer Engineering, Karaj Branch, Islamic Azad University, Karaj, Iran;

Mining frequent patterns and extracting knowledge from them is a trending field of study for industrial and business use. One of the main goals of mining frequent itemsets is to increase the accuracy of the extracted itemsets while maintaining the optimal speed and compatibility with different dataset sizes. PrePost is an algorithm for mining frequent itemsets with the least number of candidates. Although, improvements to PrePost do not respond well to large datasets and most of them do not have the optimal performance. In this research, PrePost algorithm is integrated with cloud machine learning platforms to increase the accuracy of the algorithm by reducing the number of candidates. It is expected to observe a major increase in performance and accuracy of the algorithm when applied to cloud machine learning engine.

Data Mining, Frequent Pattern Mining, PrePost Algorithm, Cloud Machine Learning Engine