Prediction of electronic cigarette and vape use among Malaysian: decision tree analysis
Introduction: The main objective of this paper is to understand the decision to use electronic cigarette and vape (ECV) and vape among Malaysian adults by assessing the perceptions and demographic variables in relations to the current status (i.e., current, former, and never use). The predictive mod...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Malaysian Medical Association
2017
|
| Subjects: | |
| Online Access: | http://irep.iium.edu.my/58874/ http://irep.iium.edu.my/58874/ http://irep.iium.edu.my/58874/1/58874_Prediction%20of%20electronic%20cigarette.pdf |
| Summary: | Introduction: The main objective of this paper is to understand the decision to use electronic cigarette and vape (ECV) and vape among Malaysian adults by assessing the perceptions and demographic variables in relations to the current status (i.e., current, former, and never use). The predictive model was developed using Induction Decision Tree (ID3) algorithm, a popular data mining technique an exploratory tool for knowledge discovery. Methods: The dataset was extracted from the National Electronic Cigarette Survey (NECS) 2016.A total of 4,288 responses were collected. The collected data was used to build and verified the model. Eight demographics variables (i.e., age, gender, race, religion, residence (urban/rural), marital, occupation and education) and twenty variables on perception of ECV were included as predictor variables. Results: By using the ID3 algorithm, it is possible to consider the relationship among variables and to identify the most informative variables for predicting the classification of the instance. It was identified that the most important variable is gender. This highlight that the decision for ECV use is significantly differ among male and female. The accuracy - i.e., percentage rate of right outcome - of the most optimum model generated in this study is 87.88%. Discussion: A number of interesting findings emerged from the ID3 model. Among others, the model indicated that young female (age < 32 years old) who perceived that ECV should be regulated than banned, and believe that ECV reduced coughing is more likely to be the current ECV user. Whereas among male, if the person is older than 44 years old, self-employed, lives in urban area, and agreed that ECV could reduce coughing, less addictive and reduced urge to smoke; he is predicted to be the current smoker. Hence, this study provides meaningful insights into understanding the different perceptions and characteristics between male and female current ECV users. |
|---|