3.4 Developed Software
We built an interactive web application on R Shiny called PDEP (Probability Density Estimation Project) which enables users to perform the same analysis that was presented in this paper for any type of univariate data. It is equipped with a wide array of features from Data Splitting, where the user can split their data into training and test datsets, as well as setting the seed and train split percentage. Furthermore, users can also plot histograms of the data with customizable binwidth and fit the data to multiple parametric models (Gaussian, Gamma, Beta & Weibull) as well as nonparametric models (KDE with ROT1 & ROT2 bandwidths and RTLLR). Moreover, the user is able to download the plots with their preferred models of fit as well as view the assessment metrics introduced in this paper which are the RMSE, MAE, MAPE, MBE, \(R^2\) and KS p-value. Analysis of Figure \ref{998686}, Table \ref{tab:train} and Table \ref{tab:test} were done through the application and Fig. \ref{711489} displays a screenshot design of the application with its current features. The application will be released once our team explores potential patents and/or journal publications.