🧠 AI for Students

how AI learns · neural networks · k-NN algorithm · types of AI · real-world impact

Inside a Neural Network

⚙️ k-NN Classifier

Place training points, choose a k value, then run the algorithm. See how k shapes the decision boundary.

k value (# neighbours)

Place dots, select a k value, then click Run k-NN.

k-Nearest Neighbours: to classify a new point, find its k closest training points and take a majority vote among their labels.

k=1 (overfit): very jagged boundary, memorises training data.
k=7 (underfit): smoother boundary, may ignore detail.

What is AI?

Artificial Intelligence (AI) is the field of building computer systems that can perform tasks that normally require human intelligence — like recognising images, understanding speech, playing games, or translating languages. Machine Learning is the most important subfield of AI today.

Why does it matter?

AI is transforming every industry — healthcare (diagnosing diseases), transport (self-driving cars), education (personalised tutoring), and science (predicting protein structures). Understanding the basics of how AI works is an essential 21st-century skill.

Key terms

Training — the process of showing a model labelled examples so it can learn patterns Model — the mathematical function that maps inputs to predictions after training Overfitting — when a model learns the training data too well and fails on new data Gradient descent — the optimisation algorithm that nudges model weights to reduce prediction error Parameters — the numbers inside a model (weights + biases) adjusted during training; GPT-4 has ~1.8 trillion

🎯 Try this challenge

In the Training Demo, place 5 blue dots in one corner and 5 pink dots in another — you'll get 100% accuracy. Now place dots randomly mixed together. Does accuracy drop? This shows why "linearly separable" data is easier for simple classifiers. Also explore what happens with k=1 vs k=5 neighbours in the description.

Continue Learning

Cyber Safety