. Author manuscript; available in PMC: 2023 Jan 1.

Published in final edited form as: Comput Speech Lang. 2021 Jul 19;71:101263. doi: 10.1016/j.csl.2021.101263

Table 3:

Truth-telling task performance using session-level aggregated features and a feature significance threshold of p <0.30.

Model	F1	Acc.	Prec.	FNR	Pos. Acc.	Neg. Acc.
DT	0.540^*	0.662^**	0.587 ^**	0.268^*	0.525	0.737
RF	0.324	0.633	0.517^**	0.328	0.250	0.840
GNB	0.378	0.536	0.370	0.354	0.393	0.613
L-SVM	0.640 ^**	0.719 ^**	0.584^**	0.165 ^**	0.725	0.718

indicates performance better than randomized bootstrap σ_t.

^**

indicates performance better than human simulation h₂. Bold values indicate the best performance in their respective columns. Pos. Acc. and Neg. Acc. indicate the accuracy for the positive and negative classes, respectively. The best performing L-SVM hyperparameters, shown in the table above, are C = 1 along with balanced class weights. The decision tree classifier used entropy as the splitting criteria, while the random forest classifier used Gini impurity.