🏅 Olympic Team Medals Predictor

This project explores a dataset of Olympic team statistics and builds a linear regression model to predict the number of medals a country will win based on features like number of athletes and prior medals.

It also evaluates the model using mean absolute error and analyzes prediction errors by country.

---

📂 Dataset

The project uses a CSV file:

which contains the following columns:

• team: Name of the team/country
• country: Country code
• year: Olympic year
• athletes: Number of athletes entered
• age: Average age of athletes
• prev_medals: Number of medals won in previous Olympics
• medals: Number of medals won in the current year

---

⚙️ What the Code Does

1. Data Exploration

   • Computes correlations between features and medals
   • Visualizes relationships with regression plots (commented out — for Jupyter)

2. Data Cleaning

   • Removes rows with missing values
   • Splits data into training and test sets (pre-2012 for training, 2012 and later for testing)

3. Model Training

   • Trains a Linear Regression model on:
     • athletes
     • prev_medals
   • Predicts number of medals for test set countries

4. Postprocessing

   • Ensures no negative predictions (clipped at 0)
   • Rounds predictions to nearest whole number

5. Evaluation

   • Computes Mean Absolute Error (MAE) on test data
   • Analyzes errors by country
   • Plots a histogram of error ratio across countries

---

📊 Key Outputs

• MAE of the model
• Per-country prediction errors
• Per-country error ratio (error divided by average medals won)
• Histogram of error ratio across countries

---

🛠️ Requirements

• Python 3.x
• pandas
• numpy
• scikit-learn
• seaborn
• matplotlib

You can install them with: pip install pandas numpy scikit-learn seaborn matplotlib

Created by dheetya