The universe is worth 64³ pixels: convolution neural network and vision transformers for cosmology

Abstract

We present a novel approach for estimating cosmological parameters, Ω m , σ8 , w 0, and one derived parameter, S 8, from 3D lightcone data of dark matter halos in redshift space covering a sky area of 40° × 40° and redshift range of 0.3 < z < 0.8, binned to 643 voxels. Using two deep learning algorithms — Convolutional Neural Network (CNN) and Vision Transformer (ViT) — we compare their performance with the standard two-point correlation (2pcf) function. Our results indicate that CNN yields the best performance, while ViT also demonstrates significant potential in predicting cosmological parameters. By combining the outcomes of Vision Transformer, Convolution Neural Network, and 2pcf, we achieved a substantial reduction in error compared to the 2pcf alone. To better understand the inner workings of the machine learning algorithms, we employed the Grad-CAM method to investigate the sources of essential information in heatmaps of the CNN and ViT. Our findings suggest that the algorithms focus on different parts of the density field and redshift depending on which parameter they are predicting. This proof-of-concept work paves the way for incorporating deep learning methods to estimate cosmological parameters from large-scale structures, potentially leading to tighter constraints and improved understanding of the Universe.

Stellar Classification with Vision Transformer and SDSS Photometric Images

“…The ViT model, based on attention mechanisms, excels in image classification tasks, demonstrating the ability to capture both global information and local details [18], making it suitable for handling multi-channel stellar photometry image data. Leveraging the ViT model for stellar image classification can enhance the utility of astronomical observation data, providing more precise and efficient data analysis tools for astronomical research [19,20].…”