A Machine-Learning Based Tool to Automate the Identification of Biological Database IDs

[Up] [Top]

Documentation for package ‘MantaID’ version 1.0.4

Help Pages

Example ID example dataset.
mi A wrapper function that executes MantaID workflow.
mi_balance_data Data balance. Most classes adopt random undersampling, while a few classes adopt smote method to oversample to obtain relatively balanced data.
mi_clean_data Reshape data and delete meaningless rows.
mi_data_attributes ID-related datasets in biomart.
mi_data_procID Processed ID data.
mi_data_rawID ID dataset for testing.
mi_filter_feat Performing feature selection in a automatic way based on correlation and feature importance.
mi_get_confusion Compute the confusion matrix for the predicted result.
mi_get_ID Get ID data from the 'Biomart' database using 'attributes'.
mi_get_ID_attr Get ID attributes from the 'Biomart' database.
mi_get_importance Plot the bar plot for feature importance.
mi_get_miss Observe the distribution of the false response of the test set.
mi_get_padlen Get max length of ID data.
mi_plot_cor Plot correlation heatmap.
mi_plot_heatmap Plot heatmap for result confusion matrix.
mi_predict_new Predict new data with a trained learner.
mi_run_bmr Compare classification models with small samples.
mi_split_col Cut the string of ID column character by character and divide it into multiple columns.
mi_split_str Split the string into individual characters and complete the character vector to the maximum length.
mi_to_numer Convert data to numeric, and for the ID column convert with fixed levels.
mi_train_BP Train a three layers neural network model.
mi_train_rg Random Forest Model Training.
mi_train_rp Classification tree model training.
mi_train_xgb Xgboost model training
mi_tune_rg Tune the Random Forest model by hyperband.
mi_tune_rp Tune the Decision Tree model by hyperband.
mi_tune_xgb Tune the Xgboost model by hyperband.
mi_unify_mod Predict with four models and unify results by the sub-model's specificity score to the four possible classes.