Univariate Anomaly Detection on NAB benchmark leaderboard · SOTA2 Research

Univariate Anomaly Detection on NAB

98.5PA-F1W1-ACAS

Evaluation Results

Method	Links
W1-ACAS2026.04		98.5	85.1	4.9	1.1	20.1	21.7
W1-ACAS2026.04		98.3	85.5	5.4	1.2	20.5	22.4
CNN*2026.04		98.2	93.7	0.8	46.8	19.4	26
W1-ACAS2026.04		98	85.8	5	1	22	22.5
conformal2026.04		97.8	88	4.8	1.3	20.5	23.2
OmniAnomaly*2026.04		97.7	92.5	6.8	50.6	20.1	27.4
conformal2026.04		97.7	87.6	5.2	1.2	20.1	22.8
conformal2026.04		97.6	88.1	5	0.8	21.7	23.1
MOMENT ZS2026.04		95.8	93.1	12.9	49	22	40.7
gaussian2026.04		94.5	81.5	14	7.8	20.3	23.5
gaussian2026.04		94.1	80	14.9	7.7	20.1	22.3
gaussian2026.04		93.5	79.2	15.4	8.5	19.5	21.8
USAD*2026.04		92.7	92.6	12.3	47.3	20	44.5
W1-ACAS2026.04		92.5	89.7	8.4	2.5	34.4	43.8
W1-ACAS2026.04		91.2	89.3	7.7	2.5	35.5	44
Sub-PCA2026.04		90.5	92.3	8.2	34.1	19.9	42.7
W1-ACAS2026.04		88.9	88.6	8.2	2.9	34.2	44.9
Conformal2026.04		87.8	89	10.7	3.8	30.8	42.9
POLY2026.04		86.3	90	13.6	23.2	8.7	32.2
Conformal2026.04		86.3	89.1	11.1	3.8	31	42
CNN2026.04		85.8	88.1	8.3	64.3	26.9	42.3
Conformal2026.04		85.1	88.5	12	4.4	31.7	44.8
KShapeAD2026.04		83.5	83.3	34.1	19.5	12.3	27.2
Sub-KNN2026.04		81	81.9	27.2	27.3	15.3	30.4
SAND2026.04		78.5	80.9	34	15.1	13	29.6
Gaussian2026.04		73.3	84.9	12.8	8.1	27	45
Gaussian2026.04		71.6	84.2	12.3	7.5	26.5	43.8
Gaussian2026.04		71.4	83.7	11.9	9	27	43.2
OmniAnomaly2026.04		67.4	85.5	20.9	57.1	16.6	42.9
MOMENT ZS2026.04		59.6	86.7	26.1	41.7	11	46.1
PCA2026.04		53.6	82.6	37.4	24.8	10	41.7
KShape2026.04		53.3	78.9	50.8	17.6	12.5	30.3
POLY2026.04		52.7	84.8	33.4	28.2	4.4	37.7
USAD2026.04		49.8	80.9	42.5	32.4	8.8	39.8
Sub-KNN2026.04		47.9	78.6	45.1	17.4	11.8	32.1
SAND2026.04		46	79	51.1	13.4	10.1	28.9