Figure 4
From: Machine learning classifies predictive kinematic features in a mouse model of neurodegeneration
Left- and right-side-only datasets yield different rankings of feature importance score, with fewer shared features in the 13-mon J20 mice. Bilateral and unilateral Feature Importance (FI) scores (%) calculated by the Random Forest feature importance method using the bilateral (a), right-side-only (b), and left-side-only (c) datasets. The 22 highest-ranked features, which made the largest contribution towards identifying genotype and age are shown. (d) A summary diagram of the 22 highest-ranked features of the left-side-only, right-side-only, and bilateral datasets indicating their common and unique features. Please see the feature definition equations for abbreviations used for each feature. Panels generated through Python 3.7 and Adobe Illustrator (2019).
