Authors
Ministry of Education, Directorate of Education Diyala, Diyala, Iraq
Abstract
Big Data demands newer optimization techniques because the growing complexity requires highly performing adaptive solutions that function efficiently. The research evaluates how the Bee Algorithm (BA) from swarm intelligence field can boost Big Data analytics and Deep Learning models. Numerous research studies demonstrate how the Bee algorithm displays strong execution in applications that involve air quality prediction with support vector regression and optimization of deep learning models along with dimensionality reduction techniques. The paper develops BI-BEE as a hybrid computational model through integration of Bee Algorithm with deep neural networks that uses dropout along with backpropagation methods. Different datasets reveal that this method effectively classifies data while also demonstrating notable overfitting resistance during tests. The research evaluates how BA can automate data classification and addresses clustering analysis and pattern extraction as well as data processing functions. The method demonstrates capability to scale and operate in parallel making it appropriate for distributed computing environments including Spark and Hadoop. Research positions BA as a strong AI tool for contemporary analytics which provides unprecedented knowledge about swarm optimization of Big Data and Deep Learning architecture.