• Ruchi Thakur

The air we breathe is one of the most important components of life on earth, and the quality of the air is becoming a serious problem for the general population. In many countries Air pollution has become a major public health risk, as a result there are many health risks which include respiratory problems, cardiovascular diseases, and even cancer. In this paper we analyses the application of two machine learning algorithms, first is time-series forecasting using the Facebook Prophet Algorithm and Regression Analysis using Support Vector Machines (SVM)—for evaluating air quality. The study consists of a comprehensive dataset provided by Himachal Pradesh Pollution Control Board (HPPCB) containing air quality indicators from various cities in Himachal Pradesh. The dataset focuses on ambient concentrations of fine particulate matter (PM10 and PM2.5), gaseous pollutants (SO2, NO2, NH3, and O3), and carbon monoxide (CO).The first step involves data preprocessing, which includes data loading, handling missing values, and ensuring date format consistency to ensure smooth integration with the forecasting and regression models. The pre-processed data is then divided into the training and testing and incorporated into the two machine learning algorithms. The accuracy of the forecasts being assessed using metrics such as mean absolute error (MAE) and root mean squared error (RMSE).On the other hand, Support Vector Regression (SVR), a version of Support Vector Machine (SVM), is being utilized to forecast Air Quality Index (AQI) values by taking into account the concentrations of various pollutants. Radial Basis function (RBF) kernel allowed to SVR for obtaining most accurate prediction. The main cause associated with air pollution are due to rising traffic volumes, extensive infrastructure projects such as tunnel construction that necessitate tree cutting, and the heavy rush of tourists to popular destinations, which increase congestion and fossil fuel emissions. Air quality prediction can assist public authorities and policy makers in developing strategies to reduce air pollution and improve public health. Highlighting their respective strengths and weaknesses in terms of accuracy, scalability and computational efficiency, the performance of both methodologies is thoroughly evaluated and compared. The findings from this comparison study offer important advice to policymakers and researchers when choosing appropriate models for evaluating and controlling air quality. This research improves the field and provides the opportunity for improved monitoring plans and accurate decision-making processes aimed at reducing the adverse impacts of air pollution on the environment and the general population, by highlighting the advantages of innovative methods for machine learning .After comparing these two machine learning algorithm Support Vector Regression (SVR) with a Radial Basis Function (RBF) kernel outperforms Prophet in terms of accuracy in predicting air quality index with the maximum coefficient of determination and less mean absolute error.

Air Quality Index, Machine Learning, Time Series Forecasting, Support Vector Regression, Facebook Prophet Algorithm

"PREDICTING AIR QUALITY INDEX AND COMPARING MODELS USING MACHINE LEARNING ALGORITHMS IN HIMACHAL PRADESH", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 7, page no.f308-f317, July-2024, Available :http://www.jetir.org/papers/JETIR2407543.pdf

"PREDICTING AIR QUALITY INDEX AND COMPARING MODELS USING MACHINE LEARNING ALGORITHMS IN HIMACHAL PRADESH", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 7, page no. ppf308-f317, July-2024, Available at : http://www.jetir.org/papers/JETIR2407543.pdf