Open Collections

Rice, Silke

University of British Columbia

The search for strongly lensed galaxies is challenging due to their rarity and the vast data volumes from upcoming surveys like Euclid. These galaxies are vital for studying dark matter and cosmology, making efficient classification methods essential. While traditional classification methods, particularly convolutional neural networks (CNNs), have proven effective, they are computationally expensive and prone to false positives. We propose a learning-to-rank algorithm with a support vector machine (SVM) classifier to prioritize candidate lensed galaxies, streamlining human inspection. Our approach outperforms CNNs in accuracy, runtime, and calibration, even with imbalanced datasets. Integrating real and simulated data ensures robust generalization to real observations. This study demonstrates that simpler models like SVMs can outperform complex deep learning architectures for this task, offering a promising alternative for classifying strongly lensed galaxies in large astronomical datasets.

Text

2025-04-24

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/90757

Statistics

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/