Abstract:—Due to the nonstationary nature, the distribution of real-world multivariate time series (MTS) changes over time, which is known as distribution drift . Most existing MTS forecasting models greatly suffer from the distribution drift and degrade the forecasting performance over time. Existing methods address distribution drift via adapting to the latest arrived data or self-correcting per the meta knowledge derived from future data. Despite their great success in MTS forecasting, these methods hardly capture the intrinsic distribution changes especially from a distributional perspective. Accordingly, we propose a novel framework temporal conditional variational autoencoder (TCVAE) to model the dynamic distributional dependencies over time between historical observations and future data in MTS and infer the dependencies as a temporal conditional distribution to leverage latent variables. Speciﬁcally, a novel temporal Hawkes attention mechanism represents temporal factors subsequently fed into feed-forward networks to estimate the prior Gaussian distribution of latent variables. The representation of temporal factors further dynamically adjusts the structures of Transformer-based encoder and decoder to distribution changes by leveraging a gated attention mechanism. Moreover, we introduce conditional continuous normalization ﬂow to transform the prior Gaussian to a complex and form-free distribution to facilitate ﬂexible inference of the temporal conditional distribution. Extensive experiments conducted on six real-world MTS datasets demonstrate the TCVAE’s superior robustness and effectiveness over the state-of-the-art MTS forecasting baselines. We further illustrate the TCVAE applicability through multifaceted case studies and visualization in real-world scenarios.

VAEneu: A New Avenue for VAE Application on Probabilistic Forecasting

Generative Probabilistic Wind Speed Forecasting: A Variational Recurrent Autoencoder Based Method

A VAE-Bayesian Deep Learning Scheme for Solar Generation Forecasting based on Dimensionality Reduction

MVG-CRPS: A Robust Loss Function for Multivariate Probabilistic Forecasting

TimeVAE: A Variational Auto-Encoder for Multivariate Time Series Generation

DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networks

History Marginalization Improves Forecasting in Variational Recurrent Neural Networks

Generative Time Series Forecasting with Diffusion, Denoise, and Disentanglement

VAE^2: Preventing Posterior Collapse of Variational Video Predictions in the Wild

Time-Causal VAE: Robust Financial Time Series Generator

Hybrid Variational Autoencoder for Time Series Forecasting

NVAE: A Deep Hierarchical Variational Autoencoder

Decision-Making with Auto-Encoding Variational Bayes

Iterative VAE as a predictive brain model for out-of-distribution generalization

Predictive variational autoencoder for learning robust representations of time-series data

Temporal Conditional VAE for Distributional Drift Adaptation in Multivariate Time Series

RVRAE: A Dynamic Factor Model Based on Variational Recurrent Autoencoder for Stock Returns Prediction

AUTO-ENCODING VARIATIONAL BAYES

VEST: Automatic Feature Engineering for Forecasting

Conditional Unscented Autoencoders for Trajectory Prediction

Probabilistic Forecasting With Modified N-BEATS Networks