Machine Learning Crash Course
summary
A whirlwind, hands-on tour that actually names models, data and outcomes: uses SVMs and scikit-learn for sign-language classification from leap motion joint coordinates, finds the keyboard UI fails and pivots to an educational game that works, forecasts Texas hourly load with k‑nearest neighbors achieving about 3% mean absolute error, mentions Facebook profit and Google causal tool, and builds a job-filter using bag‑of‑words + logistic regression that initially hits a 20% error due to class imbalance but improves by ranking with predicted probabilities.
AI-written summary · about
chapters (10)
- 0:00Start
- 1:15Machine Learning Crash Course
- 3:51Machine learning?
- 6:38Regression
- 11:55How accurate is it?
- 16:49Teach a computer sign language
- 23:18Forecast energy load
- 26:24So... what kind of ML problem is this?
- 34:51Use machine learning to find your next job
- 51:00indeed
