Air quality prediction with big data

In this thesis, the air quality index and the density of air pollutants measured at stations located throughout the province of Istanbul were estimated by means of big data and smart methods. The main problem in the thesis is the regression problem, which is one of the main problems of machine learning. Various web services and web scraping approaches have been used to obtain data from many different sources for the prediction of air quality, and to transform these data into a dataset by arranging the relevant regression problem. Afterwards, many regression methods, which are mainstream in the literature, were used on the cleaned datasets and studies were carried out to determine the hyperparameters of these methods in the best way. Obtained results are reported comparatively.