ABSTRACT
A significant challenge that is confronting our modern and rapidly industrializing society is air pollution due to its detrimental consequences to the ecosystem, its species as well as personal health. Therefore, world governments and health organizations must take measures to reduce pollutants but, unfortunately, this requires effective data and its analysis. This poses an issue since most real-world data is unorganized and contains missing values. This paper addresses this issue and several imputation methods that can be used on an air quality dataset to fill these missing values so that the data does not suffer from the loss of valuable information. The imputation methods are compared using performance measures (discussed further in the paper) to observe which method is best suited for imputing air quality datasets.
