Free Access

Table 3

Mean decrease in accuracy and node impurity of random forests trained with the heuristic set (Cols. 2 and 3) or the complete set (Cols. 5 and 6) of proper motions, magnitudes, and colour indices.

Variable Heuristic set: mean decrease in Variable Full set: mean decrease in
classification node classification node
accuracy impurity accuracy impurity

μ δ 0.0021 530.6 μ δ 0.0016 369.6

μ α 0.0015 199.4 μ α 0.0010 130.1
r − K 0.0014 156.0 K 0.0013 114.8
K 0.0024 127.1 J 0.0016 98.0
i − K 0.0008 102.0 H 0.0011 85.7
H 0.0012 80.1 r − Z 0.0008 82.0
Y − J 0.0006 77.9 Y 0.0014 70.6
J 0.0017 77.3 r − Y 0.0007 60.2
r − H 0.0005 48.7 r − H 0.0006 57.5
Y 0.0009 41.8 r − J 0.0006 55.4
Z − J 0.0002 27.3 Z 0.0011 47.0
Z 0.0007 21.6 i − Y 0.0004 44.5
i 0.0009 18.8 r − i 0.0004 43.1
r 0.0011 17.5 r − K 0.0004 42.9
i − J 0.0013 16.0 i − Z 0.0004 36.5

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.