Time series prediction with FNN-LSTM
In a recent post, we showed how an LSTM autoencoder, regularized by false nearest neighbors (FNN) loss, can be used to reconstruct the attractor of a nonlinear, chaotic dynamical system. Here, we explore how that same technique assists in prediction. Matched up with a comparable, capacity-wise, "vanilla LSTM", FNN-LSTM improves performance on a set of very different, real-world datasets, especially for the initial steps in a multi-step forecast
keras3
Machine Learning
Ai
TensorFlow
Keras
Time Series
Generative Models
"](/blog/education/my-experience-with-rstudio-instructor-training/feature-wd_hu_2c482e712ee02ba6.webp)
"](/blog/education/my-experience-with-rstudio-instructor-training/feature-wd_hu_878a15376795aeab.webp)
"](/blog/education/my-experience-with-rstudio-instructor-training/feature-wd_hu_dafb1338f7202d44.webp)
