info icon

Welcome to the IBM ICE Day 2025 Model Presentation. Test your machine Learning skills in this exciting team-based challenge.

Event Overview
General Rules & Guidelines
Challenge Breakdown
Winner Selection

Event Overview

Objective: Develop a basic machine learning model to analyze data and generate insights..
Duration: 2 Hours
Challenges: 3 (Each challenge has 3 sub-tasks)

General Rules & Guidelines

1. Team Composition

Teams must have 3-5 members incluing the Team Leader.
Teamwork and task distribution are crucial to complete the challenge within the time limit.

2. Challenge Structure

The Model Development task consists of 3 main challenges, each with 3 sub-tasks.
Teams must complete sub-tasks in sequence to unlock the next challenge.
A pre-cleaned dataset will be provided at the start of the event.

3. Scoring System

Each sub-task has different difficulty levels:
- Basic Model Setup: 50 Points
- Model Training & Accuracy Improvement: 100 Points
- Model Evaluation & Interpretation: 150 Points
The total score is calculated based on model performance, accuracy, and explanation.

4. Allowed Tools & Resources

Teams can use Python (Scikit-Learn, Pandas, NumPy, Matplotlib, Seaborn), IBM Watson Studio, Jupyter Notebook, or Google Colab.
Internet usage is allowed only for research purposes (documentation, API references, etc.).
Use of AI tools (ChatGPT, Perplexity, etc.) is not permitted for code generation.

5. Prohibited Activities

No pre-trained models or external datasets allowed. No direct copying of existing models from online sources. No dataset modifications or tampering with evaluation metrics. Any unethical behavior or rule violation may lead to disqualification.

Challenge Breakdown

Challenge Category	Task Level	Description & Requirements
Basic Model Setup	Basic (50pts)	Load the provided dataset and perform basic exploratory data analysis (EDA).
	Intermediate (75pts)	Split the dataset into training and testing sets (e.g., 80-20 split).
	Advanced (100pts)	Choose an appropriate algorithm (e.g., Linear Regression, Decision Tree, k-NN).
Model Training & Accuracy Improvement	Basic (75pts)	Train the selected model using the provided data.
	Intermediate (100pts)	Tune hyperparameters to improve model performance.
	Advanced (125pts)	Implement basic feature selection or engineering to refine the model.
Model Evaluation & Interpretation	Basic (100pts)	Evaluate the model using appropriate metrics (e.g., accuracy, precision, recall).
	Intermediate (125pts)	Visualize model performance using confusion matrices, graphs, or plots.
	Advanced (150pts)	Present findings and explain how the model can be improved.

Winner Selection

Teams with the highest total points at the end of 2 hours win. Models will be judged based on:

Correct Implementation – Is the model correctly built and trained?
Performance Metrics – Does the model provide reasonable accuracy?
Interpretability & Explanation – Can the team explain their approach clearly?
Presentation & Communication – Is the final presentation structured and well-explained?

Events: Dashboard

Events: Project