Week 20 · Data and AI

Machine learning with scikit-learn: training and testing a classifier

Last week your program learned one number. Today you use the same tools professional data scientists use — scikit-learn — to train a model that tells three species of iris flower apart from four measurements. It is a classic first project because the data is real, small and free.

Software, data and AI90 minutesAdvancedLevel 2 module

In this lesson you will

Tick them off as you read. This is a preview check-list.

Load a real dataset and split it into training and test sets
Train a k-nearest-neighbours classifier and measure its accuracy
Read a confusion matrix and explain which mistakes the model makes
Show overfitting by comparing training and test accuracy
Present a model honestly, including its limits (Project 6)

Your check-list0 of 5

Ticks here aren't saved until you sign in.

The full lesson is for enrolled students

Enrolled students get the complete notes: essential theory, key terms, a step-by-step practical walkthrough, an independent challenge, check-your-understanding questions and the take-home task.

Machine learning with scikit-learn: training and testing a classifier

In this lesson you will

Keep track of your learning

The full lesson is for enrolled students

This lesson is waiting for you