CUSTOMER LIFE TIME VALUE PREDICTION By : Kunal Pradyuman

Heroku App Link : https://customer-value-pred-kunal.herokuapp.com/

PROJECT GOAL

This project is designed to predict the CUSTOMER LIFE TIME VALUE of four wheeler insurance company using Regression Analysis with Python, FLASK, HTML, SQL

A highly comprehensive analysis with all data cleaning, exploration, visualization, feature selection, model building, evaluation and MLR assumptions validity steps explained in detail.

File Section

In Customer Lifetime Value (Exploratory Data Analysis).py

1- Data Preprocessing and some Exploratory Data Analysis to understand the data

2- Data cleaning

In Customer Lifetime Value (Feature Engineering).py

1- Data preparation: Feature Engineering and Scaling

2- Feature Selection using RFE and Model Building

3- Regression Assumptions Validation and Outlier Removal

4- Rebuilding the Model Post Outlier Removal: Feature Selection & RFE

5- Removing Multicollinearity, Model Re-evaluation and Assumptions Validation

Details of Variables [Response Variable ==> Customer Life Time Value]

OLS REGRESSION MODEL OUTPUT

Data Points vs Fitted Line

Actual Points vs Fitted Points

Technologies Used

Sample EDA VISUALIZATIONS

#############################################################################################

Gender__vs__Policy Type

From this tabulation our insight will be

         (1) When compared to Male, Female gender has taken more Policies
         (2) In all three Policies, the most prefered or Taken Policy Type is Personal Auto Policy for both Male and Female

#############################################################################################

Marital Status__vs__Policy Type

From this tabulation our insight will be

         (1) Most prefered or Taken Policy Type for all three categories is Personal Auto Policy
         (2) In all three Categories more Policy takers are Married People

#############################################################################################

Employment Status__vs__Policy Type

From this tabulation our insight will be

         (1) Most prefered or Taken Policy Type for all five categories is Personal Auto Policy
         (2) In all three Categories more Policy takers are Employed People

#############################################################################################

Employed_People vs Policy_type

From this tabulation our insight will be

         (1) Most prefered or Taken Policy Type for all three categories is Personal Auto Policy
         (2) In all three Categories more Policy takers are Employed and  Married People

#############################################################################################

UnEmployed_People vs Policy_type

From this tabulation our insight will be

         (1) Most prefered or Taken Policy Type for all three categories is Personal Auto Policy
         (2) In all three Categories more Policy takers are UnEmployed and  Single People

#############################################################################################

Employed Male People vs Policy_type

From this tabulation our insight will be

         (1) Most prefered or Taken Policy Type for all three categories is Personal Auto Policy
         (2) In all three Categories more Policy takers are Employed and Married Male People

#############################################################################################

Employed FeMale People vs Policy_type

From this tabulation our insight will be

         (1) Most prefered or Taken Policy Type for all three categories is Personal Auto Policy
         (2) In all three Categories more Policy takers are Employed and Married FeMale People

#############################################################################################

UnEmployed Male People vs Policy_type

From this tabulation our insight will be

         (1) Most prefered or Taken Policy Type for all three categories is Personal Auto Policy
         (2) In all three Categories more Policy takers are UnEmployed and Single Male People

#############################################################################################

UnEmployed FeMale People vs Policy_type

From this tabulation our insight will be

         (1) Most prefered or Taken Policy Type for all three categories is Personal Auto Policy
         (2) In all three Categories more Policy takers are UnEmployed and Single FeMale People

#############################################################################################

Married People vs Policy_type

From this tabulation our insight will be

         (1) Most prefered or Taken Policy Type for all five categories is Personal Auto Policy
         (2) In all five Categories more Policy takers are Married and Employed People

#############################################################################################

Married Male People vs Policy_type

From this tabulation our insight will be

         (1) Most prefered or Taken Policy Type for all five categories is Personal Auto Policy
         (2) In all five Categories more Policy takers are Married and Employed Male People

#############################################################################################

Married FeMale People vs Policy_type

From this tabulation our insight will be

         (1) Most prefered or Taken Policy Type for all five categories is Personal Auto Policy
         (2) In all five Categories more Policy takers are Married and Employed FeMale People

#############################################################################################

Scatter Plot Analysis