Malware Detection on EMBER benchmark leaderboard · SOTA2 Research

Malware Detection on EMBER

96.6Clean AccuracyStd. Training

Evaluation Results

Method	Links
Std. Training2021.02		96.6	73
NU-δ-MD2021.02		96.3	85
NU-δ-SHAP2021.02		96.3	85.5
NU-δ-MD2021.02		96.3	85.4
NU-δ-MDtarget2021.02		96.3	85.9
Uniform-δ2021.02		96.2	82.7
NU-δ-Mask2021.02		96.2	85.3
NU-δ-SHAP2021.02		96.2	85.2
NU-δ-MDtarget2021.02		96.2	85.4
NU-δ-Pearson2021.02		96.2	85.7
NU-δ-Pearson2021.02		96.1	85.3
Uniform-δ2021.02		96.1	83.3
NU-δ-Mask2021.02		96.1	86.7
NU-δ-MD2021.02		96	87.7
NU-δ-SHAP2021.02		96	90.3
NU-δ-Pearson2021.02		95.9	88.2
NU-δ-MD2021.02		95.9	89.9
Uniform-δ2021.02		95.8	86.3
NU-δ-SHAP2021.02		95.8	88.3
NU-δ-MDtarget2021.02		95.8	89
NU-δ-Mask2021.02		95.8	92.9
NU-δ-Pearson2021.02		95.8	90
NU-δ-Mask2021.02		95.8	91.8
NU-δ-Mask2021.02		95.7	92
NU-δ-MDtarget2021.02		95.7	90.9
NU-δ-Mask2021.02		95.7	95.8
Uniform-δ2021.02		95.6	89.3
Uniform-δ2021.02		95.6	88.4
Uniform-δ2021.02		95.5	90.17
NU-δ-SHAP2021.02		95.5	89.5
NU-δ-MD2021.02		95.4	90.54
NU-δ-MDtarget2021.02		95.4	90.7
NU-δ-SHAP2021.02		95.3	90.45
NU-δ-Pearson2021.02		95.3	90.46
NU-δ-MD2021.02		95.2	88.6
NU-δ-MDtarget2021.02		95.2	89
NU-δ-Pearson2021.02		94.9	89.7
Adv-MalBayes2023.01		89.17	—
Adv-FFNN2023.01		86.88	—
Adv-MalBayes2023.01		86.73	—
Adv-MalBayes2023.01		84.79	—
Adv-FFNN2023.01		82.44	—
Adv-FFNN2023.01		79.48	—
Adv-MalBayes2023.01		78.03	—
Adv-FFNN2023.01		64	—
Adv-MalBayes2023.01		63.06	—
Adv-MalBayes2023.01		52.63	—
Adv-FFNN2023.01		51.32	—
Adv-FFNN2023.01		42.77	—