Skip to content

Uncertainty Weighted Mean Squared Error

Home / Concepts / Uncertainty Weighted Mean Squared Error

Uncertainty Weighted Mean Squared Error

Auto-generated stub. Edit this file to add more details.

A loss function for Time-Point Processes that integrates block interval regression by weighting uncertainty with homoscedasticity.

Why It Matters

This loss function addresses the challenge of predicting discrete on-chain events by balancing event types with precise temporal (block interval) regression.

Evidence

we propose an Uncertainty Weighted Mean Squared Error (UWM) loss function, which incorporates the block interval regression term into the traditional Time-Point Process (TPP) objective function

Metadata & Links