Machine Learning System Design: A Practical Field Guide for Production & Interviews - Part 1 скачать в хорошем качестве

Machine Learning System Design: A Practical Field Guide for Production & Interviews - Part 1

Real-world ML systems are far messier than what most courses prepare you for. After going down many rabbit holes, I synthesized what matters into this Machine Learning System Design course. Here is what we'll cover: 0:00:00 Intro 0:06:56 Problem definition and goal proxies 0:13:52 Problem decomposition 0:17:16 Labels 0:21:06 Active learning 0:25:32 Stratified sampling 0:28:27 Data recency and seasonality 0:39:18 Data drifts 0:50:20 Training-serving skew 0:58:50 Time traveling 1:03:15 Importance sampling 1:06:59 Biases 1:15:35 Filtering data 1:19:50 Cold start 1:28:51 Metrics 1:49:21 Basellines 2:00:02 Slicing and dicing 2:10:35 Result diversity 2:15:58 Calibration 2:29:48 Leakage in evaluation