The improved grasshopper optimization algorithm and its applications.

Area of Science:

• Computational Intelligence
• Metaheuristic Optimization
• Machine Learning

Background:

• The Grasshopper Optimization Algorithm (GOA) is a nature-inspired metaheuristic algorithm.
• The basic GOA lacks consideration of gravity, potentially leading to slow convergence.
• Addressing this limitation is crucial for enhancing optimization performance.

Purpose of the Study:

• To improve the Grasshopper Optimization Algorithm (GOA) by introducing gravity and velocity.
• To develop an enhanced algorithm, termed Improved GOA (IGOA).
• To apply IGOA for optimizing parameters in a BP neural network for predictive modeling.

Main Methods:

• Introduced gravity force into the grasshopper position update mechanism.
• Incorporated velocity into the grasshopper position update mechanism.
• Developed IGOA with probabilistic selection of update strategies and applied it to optimize BP neural network parameters.

Main Results:

• IGOA demonstrated superior performance compared to existing algorithms on 23 benchmark functions.
• The IGOA-BPNN model achieved minimal prediction errors for Shanghai Stock Exchange Index closing prices.
• The IGOA-BPNN model also showed high accuracy in predicting the air quality index (AQI) of Taiyuan.

Conclusions:

• The proposed Improved Grasshopper Optimization Algorithm (IGOA) is effective and efficient.
• IGOA overcomes the convergence speed limitations of the basic GOA.
• The IGOA-BPNN model offers a robust approach for accurate time-series prediction tasks.