Table 3.

Performance of the five landslide machine learning models using the Wilcoxon signed-rank test (two-tailed).

No.	Pairwise Comparison	Number of Positive Differences	Number of Negative Differences	z-Value	p-Value	Significance
1	LMT vs. LR	60	50	−1.536	0.125	No
2	LMT vs. NBT	83	27	−5.590	0.000	Yes
3	LMT vs. ANN	62	46	−0.878	0.080	Yes
4	LMT vs. SVM	36	74	−3.677	0.000	Yes
5	LR vs. NBT	82	29	−5.589	0.000	Yes
6	LR vs. ANN	61	49	−0.605	0.015	Yes
7	LR vs. SVM	35	75	−4.081	0.000	Yes
8	NBT vs. ANN	36	73	−3.958	0.000	Yes
9	NBT vs. SVM	30	80	−5.711	0.000	Yes
10	ANN vs. SVM	43	67	−3.140	0.002	Yes

Note: The standard p-value is 0.05.