Intelligent and sustainable waste classification model based on multi-objective beluga whale optimization and deep...

Area of Science:

• Computer Science
• Environmental Science
• Artificial Intelligence

Background:

• Sustainable development necessitates efficient resource recycling, particularly in urban areas facing increased waste generation.
• Traditional waste management methods are insufficient for mitigating environmental harm from waste.
• Smart technologies offer advanced solutions for automated waste management.

Purpose of the Study:

• To propose an intelligent waste classification model for enhanced waste material identification.
• To improve the accuracy and efficiency of waste classification in smart cities.
• To contribute to more effective waste management strategies and sustainable practices.

Main Methods:

• Utilized the InceptionV3 deep learning architecture for waste classification.
• Employed multi-objective beluga whale optimization (MBWO) for hyperparameter tuning (dropout rate, learning rate, batch size).
• Integrated sensitivity and specificity as objective functions within MBWO for optimization.

Main Results:

• The proposed model achieved high performance on the TrashNet dataset.
• Achieved an accuracy of 97.75%, specificity of 99.55%, F1-score of 97.58%, and sensitivity of 98.88%.
• Outperformed existing state-of-the-art models in waste classification tasks.

Conclusions:

• The intelligent waste classification model effectively enhances waste material identification.
• The integration of MBWO significantly improves model accuracy and efficiency.
• The model supports more effective waste management and promotes sustainable practices in smart cities.