13. Hidden technical debt in machine learning systems
13
Sculley, David, et al. "Hidden technical debt in machine learning systems." Advances in neural information processing systems, 2015.
15. ソフトウェア開発との違い
15
https://hackernoon.com/why-is-devops-for-machine-learning-so-different-384z32f1
伝統的なソフトウェア開発フロー 機械学習の開発フロー
• User Story
• Write Code
• Submit PR
• Tests run automatically
• Review and merge
• New version builds
• Built executable deployed to
environment
• Further tests
• Promote to next environment
• More tests etc.
• PROD
• Monitor - stacktraces or error codes
• Data inputs and outputs. Preprocessd.
Large.
• Data scientist tries stuff locally with a
slice of data.
• Data scientist tries with more data as
long-running experiments.
• Collaboration – often in jupyter
notebooks & git
• Model may be pickled/serialized
• Integrate into a running app e.g. add
REST API(serving)
• Integration test with app
• Rollout and monitor performance
metrics