MOPSO for dynamic feature selection problem based big data fusion

Abstract

Optimization process occurs in many aspects and areas of everyday life. However, the big use of the internet in recent years caused a complex management of large quantities of data that are stored in many different data sources and optimization attend the domain of big data to optimize multi and dynamic data that stored in a complex dataset including all types of transactions in the data sources. So, the diversity of data stored in different data sources caused a complexity to access the information and user find a problem to present the same real world object from different sources in a clear and complementary one representation. Therefore, the high complexity of the representation of a target concept “object” that provided from different data sources, the dynamic feature selection problem based big data fusion present as a solution and a novel approach that will be applicable to solve a dynamic multi-objective optimization feature selection problem (MOOP) based on Multi-Objective Particle Swarm Optimization (MOPSO). This paper carried out on the state-of-the-art of the research done to present an overview of static and dynamic optimization in literature approach, then to define an overview of big data and to present an idea about the future work that will be able to solve the dynamic feature selection based on big data fusion with MOPSO.