Pollution and Weather Reports: Using Machine Learning for Combating Pollution in Big Cities

Area of Science:

• Environmental Science
• Computer Science
• Data Science

Background:

• Air pollution poses a significant threat to human evolution and planetary ecosystems.
• Technological advancements, including the Internet of Things (IoT), are enabling new approaches to monitor and control air pollution.
• Accessible monitoring devices have spurred innovation in environmental management.

Purpose of the Study:

• To predict air quality index (AQI) behaviors using machine learning algorithms.
• To analyze the relationship between temperature and pollution levels.
• To develop and validate a predictive model for air quality in Bucharest, Romania.

Main Methods:

• Collected air quality data from atmospheric stations in Bucharest.
• Employed three distinct machine learning algorithms for analysis.
• Studied the evolution of temperature in relation to pollution factors.
• Validated the model's performance using Root Mean Square Error (RMSE).

Main Results:

• Presented results of machine learning algorithms for predicting air quality.
• Analyzed pollutant types and their impact over two distinct periods.
• Quantified prediction errors using RMSE for each algorithm.
• Demonstrated the model's ability to forecast air quality trends.

Conclusions:

• Machine learning models can effectively predict air quality index behaviors.
• Understanding temperature-pollution dynamics is crucial for environmental management.
• The validated model provides insights into air quality evolution in urban areas.