Reducing false positives in strong lens detection through effective augmentation and ensemble learning

TL;DR

This paper demonstrates that combining high-quality, diverse training data with data augmentation and ensemble learning techniques significantly reduces false positives in CNN-based strong gravitational lens detection, improving robustness and efficiency.

Contribution

The study introduces an effective approach using data augmentation and ensemble learning to lower false positives in gravitational lens detection, with validated results on real datasets.

Findings

Achieved a false positive rate of 10^{-4}

Identified over 88% of genuine lenses

Reduced FP rate by 11-fold with minimal true positive loss

Abstract

This research studies the impact of high-quality training datasets on the performance of Convolutional Neural Networks (CNNs) in detecting strong gravitational lenses. We stress the importance of data diversity and representativeness, demonstrating how variations in sample populations influence CNN performance. In addition to the quality of training data, our results highlight the effectiveness of various techniques, such as data augmentation and ensemble learning, in reducing false positives while maintaining model completeness at an acceptable level. This enhances the robustness of gravitational lens detection models and advancing capabilities in this field. Our experiments, employing variations of DenseNet and EfficientNet, achieved a best false positive rate (FP rate) of $10^{-4}$, while successfully identifying over 88 per cent of genuine gravitational lenses in the test dataset. This represents an 11-fold reduction in the FP rate compared to the original training dataset. Notably, this substantial enhancement in the FP rate is accompanied by only a 2.3 per cent decrease in the number of true positive samples. Validated on the KiDS dataset, our findings offer insights applicable to ongoing missions, like Euclid.