Predicting maturity and identifying key factors in organic waste composting using machine learning models.

Area of Science:

• Environmental Science
• Agricultural Science
• Data Science

Background:

• Measuring the germination index (GI) in composting is currently a slow and resource-intensive process.
• Accurate GI measurement is crucial for assessing compost maturity and quality.

Purpose of the Study:

• To develop and evaluate machine learning (ML) models for predicting the germination index (GI) in composting.
• To identify key composting parameters influencing GI prediction.

Main Methods:

• Four ML models were employed: Random Forest (RF), Artificial Neural Network (ANN), Support Vector Regression (SVR), and Decision Tree (DT).
• Models were trained and validated using composting parameters to predict GI.
• SHapley additive exPlanations (SHAP) were used to determine feature importance.

Main Results:

• RF and ANN models demonstrated high predictive performance, with R² values exceeding 0.9.
• SVR (<0.6) and DT (<0.8) showed lower predictive accuracy for GI.
• Composting time, temperature, and pH were identified as significant factors influencing GI, with composting time being the most impactful.

Conclusions:

• RF and ANN are effective ML tools for accurately predicting GI in composting processes.
• This approach enables more efficient and intelligent composting management.
• The study provides a reliable method for rapid GI assessment, optimizing compost quality control.