299: Machine Learning with Paige Bailey
We've got a machine learning expert, Paige Bailey from Microsoft, on to help educate us on all the interesting things you can do with machine learning. Whether you're a huge company or a rose grower in Phoenix, there's probably something new you'll learn about machine learning in this episode.
Time Jump Links
- 1:14 Introduction of our guest
- 2:20 Introduction of our topic
- 3:40 How would you explain what machine learning is?
- 6:30 Clustering tasks vs classification tasks
- 9:58 Using a data set of dogs as an example for deep transfer learning
- 12:10 Training is a big deal - cognitive services explained
- 13:40 Accessibility benefits of machine learning.
- 15:00 Possible issues with using machine learning.
- 16:45 What about Black Mirror issues?
- 21:00 Do Facebook and YouTube have to open source their machine learning?
- 25:20 What can small e-commerce shops use if they're not Amazon?
- 27:40 What about a rose grower in Phoenix?
- 29:00 How much data do you need to make use of machine learning?
- 34:10 How do you get started with machine learning?
- 36:30 How could CodePen use the data they have?
- 39:55 What about business value of data in machine learning?
- 44:00 A warning about machine learning - future vs history
- 46:20 If the computer is wrong, how do we fix the computer?
- 49:00 Twitter using machine learning for image analysis
- 52:25 Crazy banana hats will throw off algorithms
- 56:50 How should someone get starting experimenting with machine learning?
Links
- Sarah Drasner
- NY Times: Facial Recognition Is Accurate, if You’re a White Guy
- Wired article on China’s Social Credit Score
- Weapons of Math Destruction
- Scikit Learn
- Caret Tutorial
- Tensor Flow
- Microsoft CNTK
- CodePen
- Adversarial Machine Learning
- Conferences on Machine Learning at Standuply
- Datacamp
- Andrew Ng Courses
- Azure
- AWS
- Google Cloud