Главная
Study mode:
on
1
Introduction
2
Deep Learning is not integrated with prior knowledge
3
Why is Deep Learning Fail
4
Correlation vs Causation
5
Assumption of a Stable World
6
Data Distribution Issues
7
Engineering Issues
Description:
Explore the technical challenges facing deep learning in this 13-minute video from the Neuro Symbolic Channel. Delve into an overview of critical issues in supervised machine learning, based on Gary Marcus's 2018 paper "Deep Learning: a Critical Appraisal." Examine why deep learning struggles with integrating prior knowledge, distinguishing correlation from causation, and handling unstable environments. Investigate data distribution problems and engineering hurdles that hinder progress. Access accompanying slides for a comprehensive understanding of these limitations in artificial intelligence and machine learning.

Technical Shortcomings of Deep Learning - Part 2

Neuro Symbolic
Add to list
00:00
-00:59