The relationship between environmental factors and dust accumulation by machine learning

Abstract

This study aims to explore the relationship between dust accumulation on a glass and various environmental factors including temperature, humidity, atmospheric pressure, and wind speed. The data was analyzed using Python, a popular language for data science and artificial intelligence, and regression algorithms from the scikit-learn library. The data was divided into training (80 %) and test (20 %) sets and different models were used, such as linear regression, decision tree, K-neighbor regression, random forest regression, and decision tree regression. The accuracy of the models was determined using R 2 scores, where a score of 1.0 indicates a perfect fit and negative values suggest that the model is worse than predicting the mean value. The accuracy of the selected models was calculated as a percentage by multiplying the obtained R 2 scores by 100. Graphs were used to visualise the data and determine the appropriate analysis model. The study found that the amount of dust is directly proportional to temperature and humidity. The accuracy levels of the linear models were suboptimal, leading to the use of nonlinear models like random forest regressor, decision tree regressor, and gradient boosting regressor, which showed improved performance.