Monday, December 21, 2020

On Judea Pearl: Theoretical Impediments to Machine Learning With Seven Sparks from the Causal Revolution

Recommendable! A critical short paper on the shortcomings of current machine learning research! Unfortunately, the paper only touches very briefly on some of the solutions (e.g. the three layer causal hierarchy) developed by Judea Pearl and collaborators over the past three decades. 

Pearl thinks that the present achievements of machine learning are based on not much more than statistical associations like curve-fitting.

[1801.04016] Theoretical Impediments to Machine Learning With Seven Sparks from the Causal Revolution

No comments: