Machine Learning - Opening Pandora's Box
summary
How to pragmatically introduce machine learning into products by addressing data, metrics, user experience, and system integration issues. Covers concrete practices for instrumentation, experimentation, infrastructure, planning, performance evaluation, and product-oriented ML work, based on experiences across academia, startups, consultancy, and large tech (examples include lessons learned while introducing ML at Spotify).
AI-written summary · about
chapters (14)
- 0:00Introduction
- 0:35Agenda
- 2:42Me @ QCon Rio 2015
- 3:26Me @ QCon São Paulo 2019
- 7:03Example Constraints
- 7:51Addressing Constraints
- 9:47Example Assumptions
- 12:47Addressing Assumptions
- 16:22Example Issues
- 17:53Unveiling Issues
- 20:50A/A/B Test
- 31:02Questions - cheat sheet
- 32:11Vision - cheat sheet
- 38:53Have fun!
related talks
27mA Look Inside Applied Machine Learning
27mKeys to Building Machine Learning Systems
21mHow The Application Of Machine Learning Is Changing The Way Companies Use Data
54mML Beyond Metrics: Challenges of Using ML Systems in the Real World
43mML application lifecycle
39mMachine Learning for Developers
49mUsing Machine Learning and Not Dying Trying
44mIndustrial ML – Overview of the Technologies Available for Machine Learning