README.md

In RebeccaGroh/seqbtests: Sequential Bayesian Comparison of ML Algorithm Performances

seqbtests: A Sequential Bayesian Approach for Performance Comparison of Machine Learning Algorithms.

This is a package for sequential Bayesian tests.

Getting Started

Package installation:

if (!require("devtools")) {
  install.packages("devtools")
}

devtools::install_github("RebeccaGroh/seqbtests")

Documentation

For an overview of the package functionality the package includes a vignette.

library("seqbtests") browseVignettes("seqbtests")

Once you install the package, you can access individual man pages by a call to, e.g., ?b_corr_t_test.

Package

This package provides some Bayesian approaches to compare classifiers in either single or multiple datasets. The tests [without any adaptions] in this package are modifications of tests provided by other packages, available at github, github.

Data Frame

| problem | algorithm | replication | measure_* | |:------:|:------:|:------:|:------:| | character | character | integer | numeric | | mandatory | mandatory | mandatory | mandatory |

RebeccaGroh/seqbtests documentation built on Nov. 17, 2021, 8:50 a.m.