Machine learning-based Decision Tree J48 with grey wolf optimizer for environmental pollution control.

Area of Science:

• Environmental Science
• Computer Science
• Health Science

Background:

• Industrialization and urbanization contribute to significant air pollution, posing health risks such as respiratory and cardiac problems.
• Current air quality monitoring methods are often inefficient, costly, and lack sufficient predictive capabilities.
• Accurate prediction of air quality is crucial for public health protection in metropolitan areas.

Purpose of the Study:

• To develop an efficient and accurate hybrid algorithm for predicting Air Quality Index (AQI).
• To overcome the limitations of existing air quality monitoring and prediction tools.
• To provide a cost-effective and reliable solution for assessing air quality levels.

Main Methods:

• Implementation of a hybrid algorithm combining Decision Tree J48 and Grey Wolf Optimizer (DT-GWO).
• Utilizing a dataset containing air pollutant details and AQI values, sourced from Kaggle.
• Evaluating the model's performance in predicting categorized AQI values (good, moderate, unhealthy, very unhealthy, hazardous).

Main Results:

• The proposed DT-GWO model achieved a high accuracy of 99.78% in predicting AQI.
• Individual models showed accuracies of 93.72% for Decision Tree J48 and 96.83% for Grey Wolf Optimizer.
• The DT-GWO hybrid approach demonstrated superior performance in minimizing error rates for air quality prediction.

Conclusions:

• The DT-GWO hybrid algorithm offers a highly accurate and effective method for predicting Air Quality Index (AQI).
• This model presents a significant improvement over existing methods, addressing efficiency and cost concerns in air quality monitoring.
• The findings support the use of advanced hybrid algorithms for better environmental health management and public safety.