MLSurvival: MLSurvival

In nguforche/MLSurvival: Machine Learning for Survival Analysis

Description

Train and evelaute machine learning survival and classification models for time to event data

Usage

MLSurvival(x, ...)

## Default S3 method:
MLSurvival(x, y, method, predict.times, trControl,
  parallel = FALSE, dummy.vars = TRUE, mc.cores = 2, seed = 123,
  perf = TRUE, ...)

## S3 method for class 'formula'
MLSurvival(form, dat, newdata = NULL, method,
  predict.times, trControl, parallel = FALSE, dummy.vars = TRUE,
  mc.cores = 2, seed = 123, perf = TRUE, ...)

Arguments

...	further arguments passed to caret or other methods.
method	character verctor of machine learning algorithms. Implemented algorithms glm logistic regression glmnet elastic net gbm gradient boosting machine ranger random forest svmRadial support vector machine with radial basis kernel xgbTree extreem boosting machine
predict.times	numeric vector containing the survival prediction times
trControl	list of control parameters for caret and the ranger models
parallel	run cross-validation in parallel?
dummy.vars	create dummy variables/model.matrix
mc.cores	number of cores
seed	random seed
perf	get performance metrics ?
form	survival formula
dat	data frame

Value

returns a list with items:

• model: trained survival model

• perf: performance of models at each survival prediction time: PCC, AUC, sensitivity, specificity, g-mean etc.

• perf.ave: average of perf with confidence intervals

nguforche/MLSurvival documentation built on July 28, 2019, 1:59 p.m.