Optimization of the database structure based on Machine Learning algorithms in case of increased data flow

Today, the sharp increase in the use of digital devices in the world is the reason for the creation of a large flow of data in plenty of systems. Due to the large amount of data in these systems, the process of data processing, i.e. making queries and getting the result, takes more time than traditional methods. Currently, the most popular method of real-time data processing is data processing based on a distributed computing mechanism. But one of the most important disadvantages of this method is determining the number of optimal distributions in proportion to the size of the data. Because over-distribution also has a negative effect on the efficiency of the system. In this research work, in order to solve the given problem, an approach to the optimal distribution of data flow based on Machine Learning algorithms is proposed. At the same time, during the research work, the data obtained as a result of the experiment conducted in unified computing systems based on the distributed computing mechanisms of this approach were analyzed. In addition, the efficiency indicators of 18 Machine Learning algorithms used during the ban were evaluated and the selection of the 5 most effective algorithms for the proposed approach was considered.

Перевод пока недоступен