Tobacco Smoking: Making Progress Against the Revealing Devastating Effects Using Machine Learning and Data Visualization

Tobacco smoking is the leading cause of lung cancer that links to millions of premature losses of human capital around the world, which poses an ever-increasing risk to the human being. There are two challenges in Tobacco smoking analysis: lack of latest data and unrecorded influential factors resulting from poor accessibility of census data. This paper uses machine learning and data visualization methods to analyze the relationship between Tobacco smoking and lung cancer, aiming to make suggestions on improving the smoking phenomenon. Specifically, enriching the research dataset is required, the onset of research is filling the missing data and organizing the datasets from targeted surveys and governmental or institutional websites in the preparation stage. After that, logistic regression and KNN are two models to detect cancer associated with smoking characters and gene types to confirm the correction between cancer and tobacco smoking. Finally, this paper applies Apache Echarts, Microsoft Tableau, and Python Matplotlib to realize visualizations and bring three possible solutions based on the smoking problem: increasing tax of tobacco, prohibition of cigarette advertisement, and institutional assistance for quitting smoking. The experimental results show the disastrous effect of smoking in data analysis. Even though there was a decline in the global smoking rate in past years, smoking is still unabated becoming the second risk factor that emanates in numerous deaths. Based on that, the section of visualizations displays two applicable and effective policies that is proved by official website "Our World in Data" to restrain damages of smoking.