Note this code is research code used to test a models performance on a benchmark dataset. This is not optimized or designed for production/industry applications. I simply post it here as it may help others understand my research or the topics presented here.

This repository hosts a simple notebook that showcases the Neural Stochastic Differential Equation Dynamical Variational Autoencoder (NSDE-DVAE). This was a network architecture I used in my PhD thesis to estimate the remaining useful life of turbofan engines from the popular CMAPSS dataset (link to thesis here). CMAPSS contains 4 datasets that differ based on the operating conditions of the engine and what failure modes they are subject to. The notebook showcases the NSDE-DVAE applied to the FD001 dataset but this can be changed and tested on the other datasets. The underlying theory behind the DVAE and it's application to machinery prognostics is outlined well in my other repository that goes more in-depth on how to use DVAEs for remaining useful life estimation.