ROSY Application for Selecting R Packages that Perform ROC Analysis

Abstract

The empirical ROC curve is a powerful statistical tool to evaluate the precision of tests in several fields of study. This is a twodimensional plot where the horizontal and vertical axis represent false positive and true positive fraction respectively, also referred to as 1specificity and sensitivity, where precision is evaluated through a summary index, the area under the curve (AUC). Several computer tools are used to perform this analysis one of which is the R environment, this is an open source and free to use environment that allows the creation of different packages designed to perform the same tasks in distinct ways often resulting in different customization and features often providing similar results. There is a need to explore these different packages to provide an experienced user with the simplest and most robust execution of a needed analysis. This work catalogued the different R packages capable of ROC analysis exploring their performance. A shiny web application is presented that serves as a repository allowing for efficient use of all of these packages.

Machine Learning Algorithms Based on Sampling Techniques for Raisin Grains Classification

“… Area Under Curve (AUC): An indicator used to measure and evaluate a model with a two-dimensional curve that depicts the balance between true positives and false positives (10). Table II shows the success criteria of a classification model that is determined related to the area under the curve (AUC) with a two-dimensional area [28], and the AUC value is in the range of 0.50 -1.00. Where the higher the AUC value, the better the model's performance [29].…”