🧾 Project Prompt (Assignment)

Entrena un modelo que clasifique datos

You must deliver a baseline ML classifier and a short evaluation report for stakeholders.

Train a classifier on a real dataset (Titanic, Iris, or a Kaggle tabular dataset).

Requirements

• Clean the dataset (missing values, categorical encoding)
• Engineer at least 3 features
• Train at least 2 models (baseline + improved)
• Evaluate with appropriate metrics (accuracy + at least 1 more)
• Save the trained model artifact to disk
• Write a short “model card” describing data + approach + limitations

🛠 Tools (Tech Stack)

• Python
• pandas + numpy
• scikit-learn
• matplotlib/seaborn (optional)

💡 Implementation Hints

• Use a scikit-learn Pipeline (preprocess + model)
• Fix a random seed for reproducibility
• Include a confusion matrix for classification tasks