Movatterモバイル変換

nonet

nonet is a unified solution for weighted average ensemble insupervised and unsupervised learning environment. It is a novel approachto provide weighted average ensembled predictions without using labelsfrom outcome or response variable for weight computation. In a nutshell,nonet can be used in two scenarios:

• This approach can be used in the unsupervised environment whereoutcome labels not available.
• This approach can be used to impute the missing values in thereal-time scenarios in supervised and unsupervised environment becausenonet does not require training labels to compute the weights forensemble.

Getting Started:

one of the best way to start with this project is, have a look atvignettes.Vignettes provides clear idea about how nonet can contribute to ensembledifferent models all together.

nonet also available onGithub Page

Installtion

This package can be downloaded from github using devtools:

• devtools::install_github(“GSLabDev/nonet”)

nonet uses below mentioned R version & packages:-

Requirements

• R (>= 3.5.1)

Used packages:

• caret (>= 6.0.78),
• dplyr,
• randomForest,
• ggplot2,
• rlist (>= 0.4.6.1),
• glmnet,
• tidyverse,
• e1071,
• purrr,
• pROC (>= 1.13.0),
• rlang (>= 0.2.1),

Contribution

nonet welcomes you to contribute and suggest the improvement. Kindlyraise the pull request for enhancement and raise the issue if you findany bugs.

for more details and support, one can reach out to us:

• Email: aviral.vijay@gslab.com
• (Alternate email): aviralvj@gmail.com