Towards Reliable Time Series Forecasting under Future Uncertainty: Ambiguity and Novelty Rejection Mechanisms
This work addresses forecasting reliability for real-world applications, but it appears incremental as it builds on existing rejection and detection methods.
The paper tackled the problem of unreliable time series forecasting under uncertainty by introducing a dual rejection mechanism that combines ambiguity and novelty rejection to reduce errors and adapt to data changes, enhancing model reliability in dynamic environments.
In real-world time series forecasting, uncertainty and lack of reliable evaluation pose significant challenges. Notably, forecasting errors often arise from underfitting in-distribution data and failing to handle out-of-distribution inputs. To enhance model reliability, we introduce a dual rejection mechanism combining ambiguity and novelty rejection. Ambiguity rejection, using prediction error variance, allows the model to abstain under low confidence, assessed through historical error variance analysis without future ground truth. Novelty rejection, employing Variational Autoencoders and Mahalanobis distance, detects deviations from training data. This dual approach improves forecasting reliability in dynamic environments by reducing errors and adapting to data changes, advancing reliability in complex scenarios.