Guided Evolution with Binary Discriminators for ML Program Search

TL;DR

This paper introduces a guided evolution approach using a binary discriminator to efficiently search for better machine learning programs, significantly speeding up the process across various tasks.

Contribution

It presents a novel method that employs an online-trained discriminator to guide evolutionary search in AutoML, improving speed and effectiveness.

Findings

Achieved 3.7x speedup in symbolic optimizer search

Achieved 4x speedup in RL loss function optimization

Demonstrated effectiveness across diverse ML components

Abstract

How to automatically design better machine learning programs is an open problem within AutoML. While evolution has been a popular tool to search for better ML programs, using learning itself to guide the search has been less successful and less understood on harder problems but has the promise to dramatically increase the speed and final performance of the optimization process. We propose guiding evolution with a binary discriminator, trained online to distinguish which program is better given a pair of programs. The discriminator selects better programs without having to perform a costly evaluation and thus speed up the convergence of evolution. Our method can encode a wide variety of ML components including symbolic optimizers, neural architectures, RL loss functions, and symbolic regression equations with the same directed acyclic graph representation. By combining this representation with modern GNNs and an adaptive mutation strategy, we demonstrate our method can speed up evolution across a set of diverse problems including a 3.7x speedup on the symbolic search for ML optimizers and a 4x speedup for RL loss functions.