Air Quality Index Prediction Using Simple Machine Learning Algorithms

Air pollution and its prevention are constant scientific challenges during last decades. However, they still remain huge global problems. Affecting human's respiratory and cardiovascular system, they are cause for increased mortality and increased risk for diseases for the population. Many efforts from both local and state government are done in order to understand and predict air quality index aiming improved public health. This paper is one scientific contribution towards this challenge. We compare four simple machine learning algorithms, neural network, k-nearest neighbor, support vector machines and decision tree. The air pollution database contains data for each day of 2017 from measurements stations in capital city of the Republic of Macedonia. The results are promising and it was proven that implementation of these algorithms could be very efficient in predicting air quality index