/Metis-Project-3-Churn

Using Logistic regression to interprete churn data and figure out what sort of customers are leaving a telecommunications company.

Primary LanguageJupyter Notebook

Metis-Project3: Using Logistic Regression to Interprete Churn Data and figure out what makes Customer Leaves

REDEFINED PROJECT PROPOSAL, METIS PROJECT #3

DOTUN OPASINA

ITERATION:

Initially when I began this project, I wanted to work on classification of economic growth based on different style of leadership of governments. Basically the goal was to figure out if other models of leadership can be employed by young African leaders. That proposal can be found [here] (https://github.com/Oladotun/Metis-Project-3) . Unfortunately due to insufficient data available on the impact of government styles on economic, I had to pivot quickly.

ITERATIVE PROCESS:

On that note, I have decided to change my research question to focus on using Logistic regression on Churn Data to figure out what are the prominent featues that makes a customer leave.

SCOPE:

Many companies always face the challenge whereby they begin to lose paying customers who registered for their services. This process is described as churning of the customer. I intend to provide recommendations on the type of customers leaving and what my particular company can do about it.

METHODOLOGY:

  1. Get dataSet from Kaggle
  2. Calculate Player Player Efficiency Rating (PER)
  3. Build linear regression model using current scraped data to predict salaries of players

DATA SOURCES:

TARGET

  • MVP: Find out the most important prominent feature of customers that leads to Churn.

THINGS TO CONSIDER

  • Scraping the data may take longer than expected.
  • The returned features that lead to churn may not provide any insightful reason for the customers leaving.