Abstract: The cloud computing storage approaches have been increasingly adopted by several experts from various domains. The cloud computing system interacts with various systems and these services are offered from data centres all over the world. The Multi-Agent System (MAS) includes various data agents that interact with the Cloud Data Storage (CDS) which in turn interacts with the Data Mining Agents. The data mining agents analyses data from different perspectives and summarizes it into useful information. The main process in the distributed environment for providing security and optimized access for data are: the distributed data mining provides the in-depth information of the data and provides interaction with various processing agents in parallel. The publically available security standards are introduced to incorporate cloud security goals, to avoid cloud threats. The distributed data mining systems provides improved access to the data with the analytical view of the data stored in cloud data storage. In this study, new algorithms for finding association rules and sequential pattern mining are proposed for data analysis from cloud databases. For this purpose, the data are encrypted using Hill cipher and they are analyzed and stored in encrypted form. For this purpose, a MAS is used for performing analysis effectively. The experimental results show the performance of the proposed model provides higher security and accurate analysis in comparison with other existing works.