Framework for Environment Quality Monitoring Using Radial Support Vector Regression

Abstract

<jats:p>Air contamination is the difficult issue that one must consider and is brought about by injurious gases present in the climate. For example, Carbon Dioxide, Carbon Monoxide, Sulfur Dioxide and so on. The dimensions of quality of the air shifts starting with one spot then onto the next. As indicated by WHO (World Health Organization) air contamination is the fifth significant reason for heart related diseases, hypertension, poor sustenance and tobacco smoking. Observing the measure of destructive gases over a specific zone can lessen the odds of jeopardize to individuals and caution to take prudent steps and do fundamental solutions for direct the emanation of harmful climatic gases. The present paper manages the observing of the polluted gases utilizing gas sensor which is connected with Node MCU. The levels of polluted gases sent through web to cloud stages utilizing MQTT conventions. The information is then sent to ThingSpeak cloud which can be additionally dissected from anyplace in world. Further analysis of the gas levels and estimations of future status of the air quality must be done and in order to carry out the same, high end strategies are adopted like Machine learning (ML) computation called Support Vector Regression (SVR) using Radial Basis Function (RBF), which is good in class regression analysis in the field of data prediction and significantly utilized for information anticipating. The implementation of proposed ML algorithm enhances the chances of forecasting close gas levels to actual values. And mean square error analysis gives the performance measure of the proposed model.</jats:p>