Quantitative inference of the H2 column densities from 3 mm molecular emission: case study towards Orion B

Open Access

Table 3

Statistical comparison of the performances of three regression methods on the test set (the Horsehead pillar).

Method	Hyper-parameters	Max. error^(a)	Mean error^(a)	RMSE	MSE
		dex	dex	dex	dex
Linear regression	0	0.37	0.026 ± 0.003	0.14	0.0182 ± 0.0007
Linear regression on asinh(I)	1	0.33	0.075 ± 0.002	0.11	0.0070 ± 0.0003
Random forest	3	0.26	0.040 ± 0.002	0.09	0.0060 ± 0.0002

Method	Hyper-parameters	Max. error^(a)	Mean error^(a)	RMSE	MSE
		10^dex	10^dex	10^dex	10^dex

Linear regression	0	2.34	1.062	1.38	1.043
Linear regression on asinh(I)	1	2.14	1.190	1.29	1.016
Random forest	3	1.82	1.096	1.23	1.014

Notes. ^(a) The maximum and mean errors are absolute errors on .

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.