Chaos-Embedded Multi-Objective Intelligent Optimization-Based Explainable Classification Model for Determining Cherry...

Area of Science:

• Agricultural Entomology
• Computational Biology
• Machine Learning

Background:

• The European cherry fruit fly (Rhagoletis cerasi L.) is a major pest impacting cherry production, causing significant economic losses.
• Effective pest management requires accurate assessment of infestation levels and population dynamics.

Purpose of the Study:

• To develop a novel, explainable, data-driven classification model for identifying European cherry fruit fly population classes.
• To support targeted and sustainable pest control strategies in cherry orchards.

Main Methods:

• Adapted the Chaotic Rule-based-Strength Pareto Evolutionary Algorithm 2 (CRb-SPEA2) for multi-objective optimization.
• Integrated Tent chaotic mapping to enhance population diversity and model performance.
• Utilized 10 pomological fruit characteristics to classify cherry fruit fly density without pre-defined attribute discretization.

Main Results:

• The model achieved superior classification results across all determined population classes.
• The highest accuracy rate of 82.6% was recorded for the High infestation class.
• The model demonstrated excellent sensitivity and recall values, indicating high performance.

Conclusions:

• The proposed biologically inspired, data-driven model effectively classifies cherry fruit fly populations based on fruit characteristics.
• This approach offers a transparent and interpretable tool for precision pest management in agriculture.
• The model's high accuracy supports targeted interventions, contributing to sustainable cherry production.