Abstract
Revealed by the effect of indoor pollutants on the human body, indoor air quality management is increasing. In particular, indoor smoking is one of the common sources of indoor air pollution, and its harmfulness has been well studied. Accordingly, the regulation of indoor smoking is emerging all over the world. Technical approaches are also being carried out to regulate indoor smoking, but research is focused on detection hardware. This study includes analytical and machine learning approach of cigarette detection by detecting typical gases (total volatile organic compounds, CO2 etc.) being collected from IoT sensors. In detail, data set for machine learning was built using IoT sensors, including training data set securely collected from the rotary smoking machine and test data set gained from actual indoor environment with spontaneous smokers. The prediction accuracy was evaluated with accuracy, precision, and recall. As a result, the non-linear support vector machine (SVM) model showed the best performance with 93% in accuracy and 88% in the F1 score. The supervised learning k-nearest neighbors (KNN) and multilayer perceptron (MLP) models also showed relatively fine results, but shows effectivity simplifying prediction with binary classification to improve accuracy and speed.
Highlights
In modern society, most people mainly live indoors
The non-linear support vector machine (SVM) model showed the best performance with 93% in accuracy and 88% in the F1 score
The Internet of Things (IoT) technology is used to manage and improve the cities which assert and service for the overall system This paper aims to detect pollution of indoor smoking in real time and accurately predict the pollution caused by indoor smoking and various general indoor air pollutions using
Summary
The importance of indoor air pollution has been continuously mentioned [1]. Research results, each year, 3.8 million people worldwide prematurely die from illnesses attributable to indoor air quality [2]. There are various factors such as outdoor air inflow, cooking, smoking, building materials, etc. Among them, heating, cooking and smoking were reported as typical indoor air pollution produced by human activities with purpose [3]. Cooking and heating are essential parts of our lives while indoor smoking is only the preference but adversely affects our health. According to WHO, indoor smoking exhausts fine particles 100 times higher than acceptable levels [4].
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have