=== Instructions for MATLAB files === These files are for the final app version of the UPLD framework. You need to have app designer installed for this to work. The app has not been packaged so that the code is viewable and the app can be adjusted where needed. ANY USE OF THIS APP MUST GIVE CREDIT TO THE AUTHOR - ALEX GRENYER - BY CITING THE PAPER THIS DATA IS SUPPORTING: Grenyer, A., Schwabe, O., Erkoyuncu, J. A., & Zhao, Y. (2022). Multistep prediction of dynamic uncertainty under limited data. CIRP Journal of Manufacturing Science and Technology, 37, 37–54. https://doi.org/10.1016/j.cirpj.2022.01.002 INSTRUCTIONS FOR INTIIAL USE 1. Open 'aUPLD'. This is the main app, which calls the 'loadData' and 'pol3DVectors' apps as needed. Do not run these individually. 2. Select the dataset in the top left corner. This calls the 'loadData' app. 3. Once data is loaded, train the LSTM networks. 4. Once trained, the slider can be used to adjust the train/test partition. 5. Use the tabs at the top of the app to view and edit different elements including LSTM architecture, prediction errors and spatial geometry functions. 6. Click the 'Plot 3D vectors' button to open the 3D visualisation app and toggle different elements to view. Note: The app has indicator lamps to show the status of different sections. If a lamp is red, wait for it to finish processing and turn green. If it remains red, check the MATLAB command window for any errors. Further details on the app's functionality can be found in Appendix F of the author's PhD thesis: "ADVANCED UNCERTAINTY QUANTIFICATION WITH DYNAMIC PREDICTION TECHNIQUES UNDER LIMITED DATA FOR INDUSTRIAL MAINTENANCE APPLICATIONS"