A Look Inside Applied Machine Learning
summary
Describes concrete steps for moving ML from experiments into production: build a roadmap, run proofs of concept on a governed platform, and fold data scientists into IT rather than letting them copy data to laptops. Names tools and patterns: Cloudera Data Science Workbench deployed with Cloudera Manager, container tooling like Docker/Kubernetes, and reuse of models such as LSTM reused across finance and video. Gives numbers: a churn model example at 65% predicted churn and GDPR fines of 20 million euro or 4% revenue as business risks.
AI-written summary · about
chapters (23)
- 0:00Introduction
- 1:02Mind doesnt matter
- 1:43How to grow a business
- 2:23Customer churn data
- 2:54Connecting things
- 3:28Protecting the business
- 3:51Making machine learning boring
- 4:19Why machine learning is boring
- 4:50Get away from novelty problems
- 5:50Machine learning in organizations
- 6:24Strategy
- 10:33Platform
- 11:45Vocabulary
- 13:17AI Factories
- 14:18Challenges
- 15:27Data Scientists
- 16:02Data Management
- 17:35Machine Learning
- 17:57Fast Track Knowledge Transfer
- 21:03Our Offerings
- 23:37What is Applied Machine Learning
- 24:35Interpretability
- 26:12Conclusion
related talks
21mHow The Application Of Machine Learning Is Changing The Way Companies Use Data
14mGetting Machine Learning applications into production using MLOps and CD4ML
32mMaking Models Work: How an optimized ML lifecycle can drive better ROI
27mKeys to Building Machine Learning Systems
06mMachine learning is about your data and deployment
34mScaling ML in the Traditional Enterprise
47mMachine Learning in Production: An Intro to MLOps.
28mLeading machine learning research journeys with happy (into-production) endings