
[26] Kevin Ellis - Algorithms for Learning to Induce Programs
The Thesis Review
00:00
Exploring Intelligent Systems through Discrete and Continuous Learning
This chapter explores how the combination of discrete and continuous systems can enhance machine learning models, particularly in their ability to interpret the visual and physical world. It advocates for a problem-driven approach inspired by cognitive science, aimed at mimicking human language acquisition and learning from minimal information.
Transcript
Play full episode