Skip to main content

AI / ML Lab — Iris classifier (Level 3)

Train a small classifier on the Iris dataset using Python and scikit-learn. Learn the ML pipeline: load data, train a model, evaluate, and inspect results.

Prerequisites​

  • Python 3.8+ installed
  • Optional: virtual environment
  • Dependencies: scikit-learn, numpy (see repository requirements.txt)

Files

  • notebooks/ai_ml_demo.py — a small script that trains a logistic regression on Iris and prints accuracy and a classification report.

Exercises

  1. Run the demo:

    python notebooks\ai_ml_demo.py

    Observe the reported accuracy and sample predictions.

  2. Modify the script to try a different classifier (e.g., DecisionTreeClassifier or KNeighborsClassifier). Compare the results.

  3. Introduce a simple preprocessing step: standardise features using sklearn.preprocessing.StandardScaler and see whether accuracy improves.

  4. (Optional) Plot a confusion matrix for the predictions and interpret misclassifications.

Notes for instructors

  • This lab is intentionally short and portable. It uses the small built-in Iris dataset so no external downloads are needed.