Can Machine Learning Provide Understanding? How Cosmologists Use Machine Learning to Understand Observations of the Universe

Helen Meskhidze

doi:10.1007/s10670-021-00434-5

Abstract

The increasing precision of observations of the large-scale structure of the universe has created a problem for simulators: running the simulations necessary to interpret these observations has become impractical. Simulators have thus turned to machine learning (ML) algorithms instead. Though ML decreases computational expense, one might be worried about the use of ML for scientific investigations: How can algorithms that have repeatedly been described as black-boxes deliver scientific understanding? In this paper, I investigate how cosmologists employ ML, arguing that in this context, ML algorithms should not be considered black-boxes and can deliver genuine scientific understanding. Accordingly, understanding the methodological role of ML algorithms is crucial to understanding the types of questions they are capable of, and ought to be responsible for, answering.

Highlights

The scientific value of cosmological observations, observations of the large-scale structure of the universe that are used to infer various cosmological parameters, hinges critically on the simulations used to interpret them.1 The increasing precision of the observations, has created a problem for simulators: the simulations necessary to interpret the observations have become too large and too complex for meaningful analyses to be performed on or with them
Heitmann et al consider a mock data set of 10 test cosmological models and find that emulation reproduces the nonlinear matter power spectrum to within 1% accuracy (2009, 167)
I hope to have addressed the worry about the ability for what seem like black-boxes to increase scientific understanding

Summary

Introduction

The scientific value of cosmological observations, observations of the large-scale structure of the universe that are used to infer various cosmological parameters, hinges critically on the simulations used to interpret them. The increasing precision of the observations, has created a problem for simulators: the simulations necessary to interpret the observations have become too large and too complex for meaningful analyses to be performed on or with them. The scientific value of cosmological observations, observations of the large-scale structure of the universe that are used to infer various cosmological parameters, hinges critically on the simulations used to interpret them.. The increasing precision of the observations, has created a problem for simulators: the simulations necessary to interpret the observations have become too large and too complex for meaningful analyses to be performed on or with them. The parameter spaces the simulations investigate are enormous and the simulations themselves exhibit non-linear behavior. Running these sorts of simulations has become. H. Meskhidze impractical because of the associated computational expense. Meskhidze impractical because of the associated computational expense To address this difficulty, simulators have turned to machine learning (ML) algorithms.

Methods

Findings

Discussion

Conclusion