HOLISMOKES - XI. Evaluation of supervised neural networks for strong-lens searches in ground-based imaging surveys

Open Access

Table 7

Performance of committees of networks.

	CNN			ResNet

Variations	AUROC	TPR₀	TPR₁₀	AUROC	TPR₀	TPR₁₀
Baseline	0.9557	0.0	44.4	0.9913	36.0	55.0

Diff. init.	0.9719	16.4	55.0	0.9881	15.9	52.9
Diff. splitting	0.9717	27.5	53.4	0.9725	24.9	50.8
Diff. datasets	⋯	⋯	⋯	0.9962	59.3	78.3

Notes. Committees of networks with averaging of output scores. The committee trained on different data sets is only shown for the best combination obtained with the ResNet architecture (see Section 5.10).

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.