Streamlit Dashboard for Customer Subscription Prediction

View Live Dashboard

Table of Contents

• Overview
• Dashboard Features
  • File Upload
  • Data Visualization
  • Prediction Engine
  • Results Display
  • Executive Summary
  • Customer Reports
• Deployment Guide
  • Project Structure
  • Dependencies Management
  • Path Configuration
  • Configuration Setup
  • Testing & Optimization
  • Monitoring

Overview

This dashboard implements CI/CD for model deployment and monitors performance to address model drift and related issues. It provides an interactive interface for predicting customer term deposit subscriptions based on various features.

Dashboard Features

1. Upload a CSV File

• Upload custom dataset via CSV
• Default test dataset available
• Automatic data validation

2. View Clean Test Data

• Display processed dataset
• Column overview
• Data quality checks

3. Make Predictions

• Automated prediction pipeline
• Pre-trained model integration
• Real-time processing

4. View Result Table

• Complete dataset display
• Prediction labels
• Confidence scores

5. View Executive Summary

Key metrics displayed:

• Age distribution
• Job categories
• Marital status
• Education levels
• Subscription probabilities

6. Generate Final Report for Selected Customer

Detailed customer analysis including:

• Tier Classification: Based on subscription probability
• Customer Profile:
  • Subscription probability score
  • Demographic information
  • Historical behavior
• Prediction Details:
  • Subscription status
  • Default history
  • Previous interactions

Deployment Guide

1. Project Structure

marketing_campaign_ml_prediction_dashboard/
├── .streamlit/
│   └── config.toml
├── virtual_env/
├── README.md
├── app.py
├── best_xgb.pkl
├── clean_test_data.csv
└── requirements.txt

2. Dependencies Management

# Create virtual environment
python -m venv virtual_env

# Activate environment
source virtual_env/bin/activate  # Unix
virtual_env\Scripts\activate     # Windows

# Generate requirements
pip freeze > requirements.txt

3. File Paths

# Best practices for file paths
import os
base_path = os.path.abspath(os.path.dirname(__file__))
data_path = os.path.join(base_path, 'data')

4. Configuration Files

# .streamlit/config.toml
[theme]
base="light"
primaryColor="#ad200f"
backgroundColor="#f0f7f3"
secondaryBackgroundColor="#f5970c"
textColor="#000000"
font="sans serif"

5. Testing and Optimization

# Performance optimization example
@st.cache
def load_data():
    # Expensive data loading operation
    return data

# Local testing
streamlit run app.py

6. Cloud Monitoring

• Monitor application logs
• Track performance metrics
• Debug deployment issues
• Implement error handling