Software Development for Machine Learning in Python
summary
Uses the iris example with pandas and scikit-learn to build a tiny end-to-end pipeline, persists the classifier with pickle, refactors into Model and Dataset objects, and applies dependency injection so backends swap easily. Demonstrates letting abstractions leak and adding sensible checkpoints (expose DataFrame inputs, hookable save/load) to keep fast exploration and seamless tracking. Shows a working small framework and explains tradeoffs of wider APIs versus declarative, testable production code.
AI-written summary · about
chapters (4)
- 0:00Start
- 2:20Iris classification playground
- 5:15Iris classification production
- 19:01Requirements change
related talks
45mDjango for AI: Deploying Machine Learning Models with Django with Will Vincent
26mSkynet 101 How to Keep Your Machine Learning Code From Getting Away From You
24mFrom Model to Microservice - Machine Learning at Scale
36mBuild prod ML systems with feature /training /inference pipelines
47mData Science and Machine Learning with Python
01h 27mLearning scikit learn - An Introduction to Machine Learning in Python
51mHow to Test Artificial Intelligence and Machine Learning Solutions?
30mUnit-tests for machine learning pipelines