Air pollution is a severe environmental issue these days. Rapid growth of infrastructure, industries, and vehicular density has resulted in an increase in air pollutants, including Particulate Matter (PM) concentration in the atmosphere. Exposure to PM has several harmful impacts on human health, including respiratory issues as both PM 2.5 and PM 10 are respirable. Therefore, for effective air pollution control and monitoring, this work focuses on the development of efficient prediction model for forecasting of PM 2.5 and PM 10 concentration using Machine Learning (ML) approach and Artificial Neural Networks (ANN). Prediction models were developed using ANN for PM 2.5 and PM 10 by taking the known sets of input data of 760 days from a pollution control and monitoring centre in Chandigarh city, India. This data was used to train, test and validate with different neural network tools and training algorithms. Data pre-processing and training was carried out using MATLAB software. Developed ANN models trained with Bayesian Regularization algorithm were found most efficient. Obtained accuracy was 97.33% with MSE of 25.47 and 93.60 with MSE of 230.94 for PM 2.5 and PM 10, respectively, which is very good for prediction of PM 2.5 and PM 10 concentrations in the atmosphere. The developed ANN model can be implemented by air pollution controlling and regulating authorities for forecasting concentration of PM to achieve effective air pollution control and management.