Table 1 Selected feature subsets and classification performance.

From: Quantifying ultrasonic mouse vocalizations using acoustic analysis in a supervised statistical machine learning framework

GSO

mRMR

SIMBA

RFaccuracy

duration

min freq (meanentire)

duration

duration

peak freq (stddeventire)

duration

quart 50 (start)

peak freq (stddeventire)

max freq (stddeventire)

quart 25 (meanentire)

peak freq (stddeventire)

max freq (stddeventire)

quart 50 (stddeventire)

bandw (start)

peaktopeak

quart 50 (stddeventire)

min freq (stddeventire)

peak freq (end)

max freq (stddeventire)

min freq (stddeventire)

quart 50 (minentire)

peak freq (stddeventire)

peak freq (start)

quart 50 (minentire)

quart 25 (mean)

peak freq (start)

peak freq (end)

quart 25 (mean)

max freq (minentire)

peak freq (centre)

quart 75 (start)

max freq (minentire)

quart 50 (start)

min freq (start)

entropy (mean)

quart 50 (start)

peak freq (start)

fundamental (max)

bandw (stddeventire)

peak freq (start)

peak freq (minentire)

quart 25 (mean)

entropy (maxentire)

peak freq (minentire)

min freq (minentire)

entropy (centre)

bandw (maxentire)

min freq (minentire)

max freq (start)

quart 75 (end)

quart 25 (meanentire)

max freq (start)

quart 25 (meanentire)

max freq (stddeventire)

peak ampl (end)

quart 25 (meanentire)

quart 75 (stddeventire)

quart 25 (end)

min freq (mean)

quart 75 (stddeventire)

82.0 ± 8.3

87.4 ± 6.3

88.0 ± 6.4

84.2 ± 7.4