When Flowers First Taught Computers to See

Imagine teaching a computer to recognise different species of flowers using nothing but numbers on their petals and stems. That’s exactly what happened in the 1930s with the famous Iris dataset, which kick-started our journey into data classification (grouping data into categories) and visualisation (turning numbers into pictures). In this post, we’ll see how a handful of measurements from three iris flowers led to some of the first scatter-plots and classification methods still behind AI today.

Where did this come from?

In 1936, the statistician Ronald A. Fisher published a short paper using measurements of iris flowers—sepal length, sepal width, petal length and petal width—to demonstrate a statistical technique called linear discriminant analysis. He showed that by plotting these four numbers on simple x–y charts (a scatter-plot matrix), you could draw straight lines to separate the three Iris species. Fun fact: Fisher didn’t have a computer—he did all his calculations by hand and pencil!

Where you'll see this in real life

1. Spam filters: Email services use classification algorithms to decide if a message is spam or not, often visualising patterns of words and sender details. 2. Medical diagnosis: Computers classify tumour images as benign or malignant, and doctors view colour-coded heat maps to spot problem areas. 3. Credit scoring: Banks classify loan applicants into risk categories, then use bar charts and risk curves to set interest rates. 4. Music recommendations: Streaming apps group songs into genres or “moods” by analysing audio features and show you visual playlists based on your taste.

Why it matters at school

In Stage 5 and 6 maths, you learn to sort data into categories (classification) and draw graphs (visualisation). Playing with real datasets—like the Iris data in Excel or Python—gives you a feel for how numbers turn into patterns. Next time you plot a scatter graph or colour-code exam results, remember you’re walking in Fisher’s footsteps and mastering skills at the heart of data science.