Selected article for: "forest method and random forest method"
Author: Yujia Xiang; Quan Zou; Lilin Zhao
Title: VPTMdb: a viral post-translational modification database
  • Document date: 2020_4_2
    • ID: kl99afiu_59
    Snippet: However, the combination of EGAAC, BINARY, ZSCALE and CTriad features did not significantly enhance the model's performance, which suggests that high-dimensional features may include useless features that weaken the model performance. Among all the features, considering the three evaluation values of F1-score, MCC, AUC and dimensions, the AAC combined with the ZSCALE performed best, and the sensitivity, AUC and F1-score were higher than the singl.....
    Document: However, the combination of EGAAC, BINARY, ZSCALE and CTriad features did not significantly enhance the model's performance, which suggests that high-dimensional features may include useless features that weaken the model performance. Among all the features, considering the three evaluation values of F1-score, MCC, AUC and dimensions, the AAC combined with the ZSCALE performed best, and the sensitivity, AUC and F1-score were higher than the single z-scale features. The independent test also shows that AAC combined with ZSCALE features significantly increased the AUC, F1-score, MCC, and Sn by 0.90%, 21.7%, 2.60%, and 25.0%, respectively (Supplementary Table S2 ). Now, it is important to answer two questions: (i) what is the difference between phosphorylated sites and non-phosphorylated sites and (ii) which features contribute most to the viral phosphorylation protein? To this end, we analysed the z-scale feature information between phosphorylated sites and non-phosphorylated sites. Then, we selected the most important features from the combined features with the mRMR method and using svm, random forest and naïve Bayes to perform a predictive evaluation.

    Search related documents:
    Co phrase search for related documents
    • AUC sensitivity and evaluation value: 1, 2
    • AUC sensitivity and high dimensional feature: 1
    • AUC sensitivity and important feature: 1, 2, 3, 4
    • AUC sensitivity and independent test: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11
    • AUC sensitivity and model performance: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48
    • best perform and model performance: 1
    • combined feature and feature information: 1, 2, 3
    • combined feature and high dimensional feature: 1
    • combined feature and important feature: 1
    • combined feature and model performance: 1, 2, 3, 4