Multi-fidelity emulator for large-scale 21 cm lightcone images: a few-shot transfer learning approach with generative adversarial network

TL;DR

This paper introduces a multi-fidelity, transfer learning-based GAN emulator for large-scale 21 cm lightcone images, significantly reducing computational costs while maintaining high accuracy in simulating the epoch of reionization.

Contribution

It presents a novel multi-fidelity approach using few-shot transfer learning to efficiently emulate large-scale 21 cm images with GANs, reducing computational resources needed.

Findings

Achieves percentage-level accuracy on small scales

Error increases mildly on large scales

Reduces computational cost by 10-100 times

Abstract

Emulators using machine learning techniques have emerged to efficiently generate mock data matching the large survey volume for upcoming experiments, as an alternative approach to large-scale numerical simulations. However, high-fidelity emulators have become computationally expensive as the simulation volume grows to hundreds of megaparsecs. Here, we present a {\it multi-fidelity} emulation of large-scale 21~cm lightcone images from the epoch of reionization, which is realized by applying the {\it few-shot transfer learning} to training generative adversarial networks (GAN) from small-scale to large-scale simulations. Specifically, a GAN emulator is first trained with a huge number of small-scale simulations, and then transfer-learned with only a limited number of large-scale simulations, to emulate large-scale 21~cm lightcone images. We test the precision of our transfer-learned GAN emulator in terms of representative statistics including global 21~cm brightness temperature history, 2D power spectrum, and scattering transform coefficients. We demonstrate that the lightcone images generated by the transfer-learned GAN emulator can reach the percentage level precision in most cases on small scales, and the error on large scales only increases mildly to the level of a few tens of per cent. Nevertheless, our multi-fidelity emulation technique saves a significant portion of computational resources that are mostly consumed for generating training samples for GAN. On estimate, the computational resource by training GAN completely with large-scale simulations would be one to two orders of magnitude larger than using our multi-fidelity technique. This implies that our technique allows for emulating high-fidelity, traditionally computationally prohibitive, images in an economic manner.