This repository contains the source of Experimentation with LSTM networks for Time series forecasting an app to carry out simple experiments of the use of LSTM networks in time series forecasting. It allows users to solves univariate and multivariate regression problems by testing different set of features, and comparing the effectiveness of the obtained results.
Packages: The application use the following packages: shiny, shinyalert, shinycssloaders, shinyglide, shinyjs, shinyWidgets,abind, dplyr, DT, english, future, GGally, ggplot2, htmlwidgets, jsonlite, keras, Metrics, pastecs, plotly, promises, readr, readxl, rhandsontable, rmarkdown, runner, scales, tensorflow, tools, tseries and zip. Must have packages installed to run the application.
Launch application:
shiny::runGitHub(repo = "davidrsch/LSTM-regressionexperiment", ref = "main")or try it at shinyapps.io.