2023 Challenge and Winners

Winners
First Prize Award Winner ($5000)Jack B, Massachussets, USA
Second Prize Award Winner ($3000)Leon C, Arizona, USA
Third Prize Award Winner ($1000)Riya P, Texas, USA

2023 US Applied AI Olympiad Challenge: Predicting Customer Churn

Challenge Overview

Businesses like banks face a major challenge known as Customer Churn, where customers leave their service and switch to competitors. To stay competitive, it is essential to understand the factors influencing a customer’s decision to leave and to take proactive measures to improve services.

Your Objective

  • Build a neural network-based classifier to predict whether a bank customer will leave or stay within the next 6 months.
  • Analyze the key drivers of customer churn and provide business recommendations.

Dataset Description

The dataset, attached here contains 10,000 customer records with 14 distinct features, including:

  • CustomerId: Unique identifier for each customer.
  • CreditScore: Customer’s credit history score.
  • Geography: Customer’s location.
  • Gender: Customer’s gender.
  • Age: Customer’s age.
  • Tenure: Duration of the customer’s relationship with the bank.
  • Number Of Products: Number of products purchased through the bank.
  • Balance: Customer’s account balance.
  • Has a Credit Card: Whether the customer owns a credit card (Yes/No).
  • Estimated Salary: Customer’s estimated salary.
  • isActiveMember: Whether the customer actively uses the bank’s services.
  • Exited: Target variable indicating churn (0 = No, 1 = Yes).

Challenge Deliverables

Participants are required to submit:

  1. A Google Colab notebook link (or other cloud links) showcasing:
    • Data cleaning and preprocessing steps.
    • Feature engineering techniques.
    • Model building and training process.
    • Model evaluation metrics.
    • Key insights from the model results.
  2. A Business Presentation (PDF format) including:
    • Business Overview: Problem statement and solution approach.
    • Key Findings and Insights: Highlight factors driving churn and their business implications.
    • Business Recommendations: Actionable steps for reducing churn and improving customer retention.
    • Potential Benefits: Explain how implementing your solution will create value for the business.

Important

  • Avoid copying code from the notebook into the presentation.
  • Focus on insights and their impact on business decisions.
  • Keep the content simple and engaging for a non-technical audience like a Data Science Lead.

Submission Guidelines

  • Challenge Opens: November 15, 2023
  • Submission Deadline: November 31, 2023
  • Submit your deliverables via the provided submission portal.

Required Files

  • Google Colab notebook or cloud link containing your code.
  • A PDF version of your business presentation.

Evaluation Criteria

Submissions will be judged based on:

  • Model Accuracy: The effectiveness of your classifier in predicting churn.
  • Insights and Analysis: Clarity and depth of your findings.
  • Business Recommendations: Relevance and feasibility of your proposed solutions.
  • Presentation Quality: Clear, concise, and impactful storytelling tailored to a business audience.

Why Participate?

  • Gain hands-on experience with real-world datasets.
  • Develop and showcase your skills in neural networks, data analysis, and business strategy.
  • Compete with peers for recognition in the field of AI and data science.

This is your opportunity to demonstrate your ability to bridge technical expertise with strategic thinking. Take the challenge and make an impact—because the future of AI starts here.