Skip to contents

Train, predict, and evaluate classification models

Source: R/splendid_model.R
splendid_model.Rd

Train, predict, and evaluate classification models

Usage

splendid_model(
  data,
  class,
  algorithms = NULL,
  n = 1,
  seed_boot = NULL,
  seed_samp = NULL,
  seed_alg = NULL,
  convert = FALSE,
  rfe = FALSE,
  ova = FALSE,
  standardize = FALSE,
  sampling = c("none", "up", "down", "smote"),
  stratify = FALSE,
  plus = TRUE,
  threshold = 0,
  trees = 100,
  tune = FALSE,
  vi = FALSE
)

Arguments

data

data frame with rows as samples, columns as features

class

true/reference class vector used for supervised learning

algorithms

character vector of algorithms to use for supervised learning. See Algorithms section for possible options. By default, this argument is NULL, in which case all algorithms are used.

n

number of bootstrap replicates to generate

seed_boot

random seed used for reproducibility in bootstrapping training sets for model generation

seed_samp

random seed used for reproducibility in subsampling training sets for model generation

seed_alg

random seed used for reproducibility when running algorithms with an intrinsic random element (random forests)

convert

logical; if TRUE, converts all categorical variables in data to dummy variables. Certain algorithms only work with such limitations (e.g. LDA).

rfe

logical; if TRUE, run Recursive Feature Elimination as a feature selection method for "lda", "rf", and "svm" algorithms.

ova

logical; if TRUE, a One-Vs-All classification approach is performed for every algorithm in algorithms. The relevant results are prefixed with the string ova_.

standardize

logical; if TRUE, the training sets are standardized on features to have mean zero and unit variance. The test sets are standardized using the vectors of centers and standard deviations used in corresponding training sets.

sampling

the default is "none", in which no subsampling is performed. Other options include "up" (Up-sampling the minority class), "down" (Down-sampling the majority class), and "smote" (synthetic points for the minority class and down-sampling the majority class). Subsampling is only applicable to the training set.

stratify

logical; if TRUE, the bootstrap resampling is performed within each strata of class to ensure the bootstrap sample contains the same proportions of each strata as the original data.

plus

logical; if TRUE (default), the .632+ estimator is calculated. Otherwise, the .632 estimator is calculated.

threshold

a number between 0 and 1 indicating the lowest maximum class probability below which a sample will be unclassified.

trees

number of trees to use in "rf"

tune

logical; if TRUE, algorithms with hyperparameters are tuned

vi

logical; if TRUE, model-based variable importance scores are returned for each algorithm if available. Otherwise, SHAP-based VI scores are calculated.

Algorithms

The classification algorithms currently supported are:

  • Prediction Analysis for Microarrays ("pam")

  • Support Vector Machines ("svm")

  • Random Forests ("rf")

  • Linear Discriminant Analysis ("lda")

  • Shrinkage Linear Discriminant Analysis ("slda")

  • Shrinkage Diagonal Discriminant Analysis ("sdda")

  • Multinomial Logistic Regression using

    • Generalized Linear Model with no penalization ("mlr_glm")

    • GLM with LASSO penalty ("mlr_lasso")

    • GLM with ridge penalty ("mlr_ridge")

    • GLM with elastic net penalty ("mlr_enet")

    • Neural Networks ("mlr_nnet")

  • Neural Networks ("nnet")

  • Naive Bayes ("nbayes")

  • AdaBoost.M1 ("adaboost_m1")

  • Extreme Gradient Boosting ("xgboost")

  • K-Nearest Neighbours ("knn")

Examples

data(hgsc)
class <- attr(hgsc, "class.true")
sl_result <- splendid_model(hgsc, class, n = 1, algorithms = "xgboost")