In the 4th to 6th week of the Metis Data Science Bootcamp, every participation is required to choose a Classification project and finally create a Flask App, hosted on Heroku, to demostrate the model's prediction. In my project I have taken a survey data set from Kaggle on Airline Passenger Satisfaction, around which I built a classification model to identify the critical factors leading to satisfaction.
Through this project, I engineered a highly precise model of 99% Precision, by tuning the probability threshold and also the hyperparameters of various classification models. I employ skills such as exploratory data analysis, feature selection and cross-validation with models as such k-Nearest Neighbors, Logistic Regression, Gaussian Naive Bayes, Decision Trees and Random Forest. Finally I applied the highly precise model to a business problem to demostrate potential use case.
Flask App: https://flight-satisfaction-prediction.herokuapp.com/
Medium Blog: https://medium.com/@tanpengshi/predicting-satisfaction-of-airline-passengers-with-classification-76f1516e1d16