The Galaxy Assembly and Interaction Neural Networks (GAINN) for high-redshift JWST observations

Abstract

We present the Galaxy Assembly and Interaction Neural Networks (GAINN), a series of artificial neural networks for predicting the redshift, stellar mass, halo mass, and mass-weighted age of simulated galaxies based on JWST photometry. Our goal is to determine the best neural network for predicting these variables at $11.5 < z < 15$. The parameters of the optimal neural network can then be used to estimate these variables for real, observed galaxies. The inputs of the neural networks are JWST filter magnitudes of a subset of five broadband filters (F150W, F200W, F277W, F356W, and F444W) and two medium-band filters (F162M and F182M). We compare the performance of the neural networks using different combinations of these filters, as well as different activation functions and numbers of layers. The best neural network predicted redshift with normalized root mean squared error NRMS = $0.009_{-0.002}^{+0.003}$, stellar mass with RMS = $0.073_{-0.008}^{+0.017}$, halo mass with MSE = $ 0.022_{-0.004}^{+0.006}$, and mass-weighted age with RMS = $10.866_{-1.410}^{+3.189}$. We also test the performance of GAINN on real data from MACS0647-JD, an object observed by JWST. Predictions from GAINN for the first projection of the object (JD1) have mean absolute errors $\langle \Delta z \rangle <0.00228$, which is significantly smaller than with template-fitting methods. We find that the optimal filter combination is F277W, F356W, F162M, and F182M when considering both theoretical accuracy and observational resources from JWST.