2022年1月28日 · Statistical properties such as mean and variance often change over time in time series, i.e., time-series data suffer from a distribution shift problem. This change in temporal distribution is one of the main challenges that prevent accurate time-series forecasting. To address this issue, we propose a simple yet effective normalization method called reversible …

s, i.e., time-series data suffer from a distribution shift problem. This change in temporal distribution is one of the main challenges that prevent accurate time-series forecasting. To address this issue, we propose a simple yet effective nor-malization method called reversible instance normalization (RevIN), a generally-applicable normalization-and-denormalization method with learnable …

2024年1月16日 · The recent boom of linear forecasting models questions the ongoing passion for architectural modifications of Transformer-based forecasters. These forecasters leverage Transformers to model the global dependencies over temporal tokens of time series, with each token formed by multiple variates of the same timestamp. However, Transformers are …

RevIN [11] was the first work to apply reversible normalization for time series forecasting, which assumes that history and future sequences share the same distribution.

RevIN Identity Y 2022), re extractor and do not update its parameters in the training phase. It is worth noting that the RevIN significantly improves the forecasting accuracy of these approaches. Thus, comparing one method with others wh ch do not use RevIN may lead to unfair results due to its advantage. With the aid of RevIN, even a simple li

Their work served to show that linear models are comparable, and sometimes better, than complex transformer architectures. Reversible instance normalisation (Kim et al., 2021) (RevIn) is a feature normalisation technique that typically improves time series forecasting.

2024年1月16日 · In this paper, we introduce FITS, a lightweight yet powerful model for time series analysis. Unlike existing models that directly process raw time-domain data, FITS operates on the principle that...

Welcome to the OpenReview homepage for ICLR 2022 Conference

2023年12月31日 · Our methodology capitalizes on a sparse vector quantization technique coupled with Reverse Instance Normalization (RevIN) to reduce noise impact and capture sufficient statistics for forecasting, serving as an alternative to the Feed-Forward layer (FFN) in the transformer architecture.

2023年9月21日 · Recently, multivariate time series (MTS) forecasting techniques have seen rapid development and widespread applications across various fields. Transformer-based and GNN-based methods have shown...