Python Idle
The postal facility is ready to go live. They need a production-grade digit recognition system — not a prototype. You must build the full pipeline from scratch: load raw images, preprocess them, train a classifier that exceeds 95% accuracy, and produce a confusion matrix so the operations team can see exactly where the model struggles. No hand-holding this time. Every import, every step — it's on you.
~30 minproject
Loading Python runtime...
Goals: 4 tests
should apply preprocessing to the data
should achieve model accuracy above 0.95
should generate a confusion matrix
Python loading...