AutoML-Zero: Evolving Code That Learns

They use evolutionary search [0] to create programs for image classification, specifically 'binary classification tasks extracted from CIFAR-10'. They do it from scratch, though they use a pytorch-ish programming language with differentiation baked in.

Looks like a nice summer intern project.

[0] https://en.wikipedia.org/wiki/Genetic_programming

"AutoML-Zero: Evolving Machine Learning Algorithms From Scratch" (2020) https://arxiv.org/abs/2003.03384 https://scholar.google.com/scholar?cluster=11748751662887361...

How does this compare to MOSES (OpenCog/asmoses) or PLN? https://github.com/opencog/asmoses https://scholar.google.com/scholar?hl=en&as_sdt=0%2C43&q=%22... (2007)

Pretty exciting stuff! The Github repo is here: https://github.com/google-research/google-research/tree/mast...

Is this symbolic AI and/or a (deep learning) neural network?