Selected article for: "testing set and training set"

Author: Xueyan Mei; Hao-Chih Lee; Kaiyue Diao; Mingqian Huang; Bin Lin; Chenyu Liu; Zongyu Xie; Yixuan Ma; Philip M. Robson; Michael Chung; Adam Bernheim; Venkatesh Mani; Claudia Calcagno; Kunwei Li; Shaolin Li; Hong Shan; Jian Lv; Tongtong Zhao; Junli Xia; Qihua Long; Sharon Steinberger; Adam Jacobi; Timothy Deyer; Marta Luksza; Fang Liu; Brent P. Little; Zahi A. Fayad; Yang Yang
Title: Artificial intelligence for rapid identification of the coronavirus disease 2019 (COVID-19)
  • Document date: 2020_4_17
  • ID: 79tozwzq_70
    Snippet: We developed support vector machine, random forest and multi-layer perceptron classifiers based on patients' age, sex, exposure history, symptoms (present or absent of fever, cough and/or sputum), white blood cell counts, neutrophil counts, neutrophil percentage, lymphocyte counts and lymphocyte percentage. We fine-tuned the hyperparameters of each classifier on the training set and tuning set, and evaluated the best model on the testing set. For.....
    Document: We developed support vector machine, random forest and multi-layer perceptron classifiers based on patients' age, sex, exposure history, symptoms (present or absent of fever, cough and/or sputum), white blood cell counts, neutrophil counts, neutrophil percentage, lymphocyte counts and lymphocyte percentage. We fine-tuned the hyperparameters of each classifier on the training set and tuning set, and evaluated the best model on the testing set. For the support vector machine classifier, we assessed the "C", and kernel. For the random forest classifier, the number of estimators was tuned. For multi-layer perceptron, we assessed the number of layers and the number of hidden nodes in each layer. After the hyperparameter optimization, a 3-layer MLP model with 64 nodes in each layer was selected because of the highest AUC score on the tuning set The MLP model was selected because of the highest AUC score on the tuning set (Extended Fig. 3) . We used the Scikit-learn 32 package to fit and evaluate these models.

    Search related documents:
    Co phrase search for related documents