Prediction of the compression ratio for municipal solid waste using decision tree

Area of Science:

• Environmental Engineering
• Waste Management Science
• Geotechnical Engineering

Background:

• The compression ratio of municipal solid waste (MSW) is critical for landfill design and settlement evaluation.
• Existing models for estimating MSW compression ratio are insufficient.
• Accurate prediction of waste compression is needed for effective landfill management.

Purpose of the Study:

• To develop a practical model for predicting the compression ratio of municipal solid waste (MSW).
• To utilize a decision tree method for estimating waste compression ratio (C'c).
• To establish a relationship between C'c and key waste properties.

Main Methods:

• Employed Quinlan's M5 algorithm to construct a decision tree model.
• Utilized a literature-based database of MSW properties.
• Incorporated waste composition factors like dry density, water content, and biodegradable organic waste percentage.

Main Results:

• Developed a decision tree model capable of predicting MSW compression ratio (C'c).
• The model effectively relates C'c to waste characteristics such as dry density and water content.
• Statistical evaluation confirmed the model's accuracy and reliability.

Conclusions:

• The developed decision tree model provides an effective method for evaluating MSW compression ratio.
• This model can significantly aid in landfill design and waste settlement assessments.
• The study addresses the need for a practical and reliable tool in waste management engineering.