Non-mainstream ML approaches

Statistical analysis of DNN layers:

• https://calculatedcontent.com/2019/12/03/towards-a-new-theory-of-learning-statistical-mechanics-of-deep-neural-networks/
• https://github.com/CalculatedContent/WeightWatcher

Theory of Over parameterized ML: http://topml.rice.edu/

This HN comment with some context on what modern DL techniques miss out on.

The field of ILP - part of Symbolic AI:

• https://en.wikipedia.org/wiki/Inductive_logic_programming
• https://arxiv.org/pdf/2008.07912.pdf
• Meta Inductive Learning: https://www.doc.ic.ac.uk/~shm/Papers/rulemlabs.pdf

ART: Adaptive Resonance Theory https://en.wikipedia.org/wiki/Adaptive_resonance_theory