This project is created by 4 students: Vu Truong, Quang Huynh, Anh Ngo, Thanh Nguyen at SP Jain School of Global Management for the purpose of learning fundamental Machine Learning. This is our first project about ML, which we hope it would be an important premise for us to build more larger projects in the future.

Implementing Machine Learning models and statistical methods to predict the churn target based on 10,000 sample of customer information. Using Python and its library (NumPy, Pandas, Matplotlib, Scikit - learn) to evaluate the most appropriate model, which has accuracy of above 80%. Generating a report to communicate and interpret the results to business users

OUTLINE

I) Data Exploration: (Thanh)

Import libraries and dataset

General description

Correlation matrix

Exploratory Data Analysis (EDA)

II) Data Cleaning and Preprocessing: (Vu)

Data Cleaning

Preprocessing

Dealing with outliers

III) Testing multiple models (Anh)

Splitting the dataset

Building and testing model

ROC curve for comparison

Precision – Recall curve

Choosing models

IV) Fit models (Thanh)

A – Linear Discriminant Analysis

B – Logistic Regression

C – Gaussian Naïve Bayes

D – K-Nearest Neighbors

V) Conclusion (Quang)

VI) Additional Information

LDA Reduced Model (Quang)

Decision Tree Model (Quang)

Defining threshold (Vu)