Multi-Variate Time Series Forecasting on Variable Subsets

by   Jatin Chauhan, et al.
Indian Institute Of Technology, Madras

We formulate a new inference task in the domain of multivariate time series forecasting (MTSF), called Variable Subset Forecast (VSF), where only a small subset of the variables is available during inference. Variables are absent during inference because of long-term data loss (eg. sensor failures) or high -> low-resource domain shift between train / test. To the best of our knowledge, robustness of MTSF models in presence of such failures, has not been studied in the literature. Through extensive evaluation, we first show that the performance of state of the art methods degrade significantly in the VSF setting. We propose a non-parametric, wrapper technique that can be applied on top any existing forecast models. Through systematic experiments across 4 datasets and 5 forecast models, we show that our technique is able to recover close to 95% performance of the models even when only 15% of the original variables are present.


page 1

page 2

page 3

page 4


Forecasting in multivariate irregularly sampled time series with missing values

Sparse and irregularly sampled multivariate time series are common in cl...

Two-Stage Framework for Seasonal Time Series Forecasting

Seasonal time series Forecasting remains a challenging problem due to th...

FreDo: Frequency Domain-based Long-Term Time Series Forecasting

The ability to forecast far into the future is highly beneficial to many...

Forecast Evaluation for Data Scientists: Common Pitfalls and Best Practices

Machine Learning (ML) and Deep Learning (DL) methods are increasingly re...

Please sign up or login with your details

Forgot password? Click here to reset