The Exponentially Weighted Moving Average

461 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Why Can GPT Learn In-Context? Language Models Secretly Perform Gradient Descent as Meta-Optimizers (2023, methodology): “By analogy with momentum SGD that averages gradients among timestamps, we try to apply Exponential Moving Average (EMA) (Hunter, 1986) to the attention values to build momentum-based attention: where η is a hyper-parameter, and v i is the i -th attention value vector.”
Applications of fault detection and diagnosis methods in nuclear power plants: A review (2011, methodology): “Signal-basedmethods in time domain include control charts (Montgomery, 2005) (Hunt, 1986), AR model fitting for step response estimation (Hashemian et al.”