A Hybrid Multi-population Optimization Algorithm for Global Optimization and Its Application on Stock Market Prediction

Meta-heuristic algorithms are formulated and implemented based on events and occurrences in nature. By creating shortcuts throughout the problem space, they try to find optimal solutions in a reasonable amount of time. Each meta-heuristic algorithm has various advantages and disadvantages due to its unique characteristics. Therefore, hybridizing or integrating special mechanisms can significantly improve their weaknesses. This paper uses a multi-population hybrid of the Slime Mould Algorithm and the Artificial Ecosystem-based Optimization to create a powerful and efficient algorithm. Also, the Quantum Rotation Gate mechanism has been used to maximize the variety of production solutions in the proposed method. Fifty-two standard benchmarks have been used to evaluate the proposed method. The obtained results indicate the excellent and significant performance of the proposed method. However, we have enhanced the Extreme Learning Machine classifier using the proposed method to forecast the stock market. Five stock market indices based on real stock market prices have been used for evaluation.