Skip to main content
ML Quest
Python Idle

Beat 95% accuracy on handwritten digit recognition. Whatever it takes. The sklearn digits dataset has 1,797 images of handwritten numbers (0-9), each represented as an 8x8 pixel grid. Your challenge: pick the right algorithm, tune it if you must, and break through the 95% barrier. No external data allowed — just skill.

~10 minbuild challenge
Loading Python runtime...
Goals: 2 tests
should achieve accuracy above 0.95
should have a trained model with .predict method
Python loading...