What is regression in AI? (Explained for kids)

Updated May 8, 2026 · 260 words

Regression is when AI predicts a number instead of a category. "What''s tomorrow''s temperature?" is regression. "How much will this house sell for?" is regression. The output isn''t a label — it''s a value on a continuous scale.

How to explain it to a 7-year-old

🧒 "Classification picks which group something belongs to. Regression guesses how much — like guessing how heavy your backpack is, or how tall your friend will be next year."

How to explain it to a 14-year-old

🎒 "Regression maps inputs to a continuous numerical output. Linear regression draws a line through points; modern regression uses neural networks for non-linear relationships. Loss is typically squared error (how far off the prediction was)."

Real-world examples

🌡️ Weather temperature forecasts
🏠 Zillow''s house-price estimates
📈 Stock price predictions (with caveats)
🚗 Estimated arrival time in Google Maps
🎬 Movie box-office predictions

Where this comes up in Chippu

Band C (c2-1) introduces regression through a "guess the height of the next plant" activity.

Related terms

Frequently asked questions

Is regression in AI the same as regression in statistics?

Same idea, different scale. Statistical regression usually fits a single equation; AI regression often uses a neural network with millions of parameters but solves the same kind of problem — predict a number from inputs.

Is weather prediction regression?

Yes — temperature forecasting is a classic regression problem. 'Will it rain' is classification (yes/no); 'how many millimeters of rain' is regression.