Create a Web App For Predicting Customer Churn using Streamlit

It's much easier to keep an existing customer than to acquire a new one. That's why churn prediction and retention management is a strategically important data science use case. In this project you'll learn how to build your own churn model and app.

Buy $99.00

Course curriculum

1. Kickoff
2. Client Conversation: Classification for Churn
1. Import Data from a CSV
1. Split Data to Prevent Data Snooping Bias
1. Client Conversation: Anything different when it comes to EDA for Classification?
2. Birds Eye View
3. Faster EDA with Pandas Profiling
4. Inspect Target Classes
5. Inspect State, Area Code and Promotions Offered
6. Inspect VMail mesaages and Customer Service Calls
7. Analyse Customer Service Calls Given it's Skewed
8. Look into NumericNumeric Correlation
9. Plot Churn vs Numeric
10. Plot Categorical Relationships to Churn
11. Relationship Between Reamining Term and NPS Rating
1. Double Check were Operating with a Clean Slate
2. Handling Missing Area Code
3. Working Missing Voice Mail Plan
4. Imputing Missing Eve Minutes
5. What to do with a Missing Target Value
1. Building a Ratio for Correlated Predictors
2. Applying Log Transforms to Skewed Customer Service Calls
3. Feature Engineering Unhappy Customers
4. Creating Target and Feature Values
5. Client Conversation: What are we doing about imbalanced classes?
6. Dealing with Imbalanced Classes using IM
1. Client Conversation: Building Classification Models
2. Build Machine Learning Pipelines
3. Creating Hyperparameter Grids
4. Train Models with GridSearchCV
1. Client Conversation: How do we check if this is working?
2. Applying Transformations to Test Data
3. Evaluate Performance Metrics on Scores
4. Understanding Performance Metrics including the Confusion Matrix
5. Making Predictions on Test Data
1. Save and Serialize the Models and Encoder
2. Create a Model Schema in JSON Format
1. Reload Models and Encoder
2. Transform Test Sample and Predict
1. Client Conversation: How exactly are we building an app for this?
2. Setup Utilities File
3. My First Streamlit App
4. Import Dependencies
5. Load JSON Schema
6. Setup Streamlit Sidebar for ML Input
7. Load Model and Encoder and Handle Mean Eve Mins
8. Build Prediction Capability
9. Handling and Visualizing Historical Data
10. Wrap Up

About this course

$99.00
49 lessons
6 hours of video content

Discover your potential, starting today

Enroll today

Create a Web App For Predicting Customer Churn using Streamlit

Course curriculum

Kickoff

Importing Data

Split Data into Training and Testing

Exploring Why Customers Might be Leaving

Data Preprocessing

Feature Engineering for Churn

Building Classification Models

Evaluation for Classification and Prediction

Deployment

Test Scoring

Creating a Customer Churn Streamlit App

About this course

Discover your potential, starting today