Super Data Science: ML & AI Podcast with Jon Krohn 469: Learning Deep Learning Together
May 11, 2021
Konrad Körding, a full professor at the University of Pennsylvania, merges biological neuroscience with artificial intelligence and leads the Neuromatch Academy. In this discussion, he emphasizes the significance of causality in deep learning and explores the future of artificial general intelligence. He shares insights on innovative educational approaches during the pandemic and the high completion rates of online courses. Körding also highlights skills essential for aspiring data scientists and discusses the role of interdisciplinary collaboration in advancing AI research.
AI Snips
Chapters
Books
Transcript
Episode notes
High Completion Rates
- Neuromatch Academy boasts an 86% completion rate compared to 5-15% for typical MOOCs.
- The interactive format and supportive groups help students stay engaged and motivated.
Time Commitment
- Dedicate three continuous weeks to Neuromatch, focusing on mornings for learning and afternoons for projects.
- Family commitments are manageable, but it's difficult to combine with a full-time job.
CNN Efficiency
- Convolutional Neural Networks (CNNs) use fewer parameters for tasks like image recognition where elements can be shifted around.
- This efficiency makes them very useful for data science applications.
