Practical AI AI code that facilitates good science
Jul 19, 2019
Joel Grus, author of 'Data Science from Scratch' and senior research engineer at the Allen Institute for AI, chats about his journey into AI and data science. He shares insights on the importance of coding best practices and how tools like Jupyter Notebooks can enhance collaboration. The conversation also delves into the exciting AI projects at AI2, such as Aristo and Mosaic, and the role of open-source initiatives like AllenNLP in advancing research. Plus, get tips for writing clean, maintainable code in machine learning!
AI Snips
Chapters
Books
Transcript
Episode notes
Career Transition
- Joel Grus transitioned from quantitative finance to data science after his hedge fund failed.
- He leveraged his math background and learned programming on the job at Faircast, a travel startup.
Data Science from Scratch
- Simply using machine learning libraries without understanding can lead to mistakes.
- Grus emphasizes the importance of understanding underlying principles, similar to proving theorems in math.
Learning Progression
- Start with Python, then linear algebra using lists.
- Build up to statistics, probability, data handling, and then machine learning concepts.
