In the current generation, it has been studied that smoking addiction among the youths is increased exponentially. Since there is a lot of awareness among the people about tobacco use but the youths are exposed to a broader spectrum of the different types of nicotine products like E-cigarettes, or hookah or water pipes, or conventional cigarettes, or dissolvable tobacco, and many more products. Our study aims are to identify some of the demographic factors like age, gender, sex, etc. to predict smoking addiction among the youths of our society. To predict e-cigarette addiction among youths, we considered the wings of Artificial Intelligence (AI) like Machine Learning (ML), and Deep Learning (DL) for better understanding and finding the relationship among the various features. During the data preprocessing we used the elastic net regression technique for feature selection and K-Nearest neighbor for making the accurate predictions. The Hybrid Prediction model was built by the Elastic net regression technique and KNN. It is observed that Elastic net finds a better selection of significant features. The outcome of the suggested pipeline provides high performance on the selection of significant features for the prediction model and provides better accuracy.