Glossary
Adaptive Alpha
A dynamic smoothing factor in exponential moving average (EMA) scoring that adjusts based on transaction volume and recency.
What is Adaptive Alpha?
In trust scoring systems, adaptive alpha allows recent performance to influence scores more heavily for agents with consistent activity, while preventing volatile swings for agents with sparse transactions. The alpha value typically ranges from 0 to 1, where higher values give more weight to recent observations.
This approach balances responsiveness to changing behavior with stability in score representation. Signet uses adaptive alpha to ensure scores reflect current agent reliability without overreacting to isolated incidents or underweighting sustained patterns.
Example
An agent with daily transactions might have alpha=0.3, making scores responsive to trends, while an agent with monthly activity uses alpha=0.1 to prevent single events from dominating their reputation.
How Signet addresses this
Signet's scoring methodology employs adaptive alpha in the Reliability Score calculation, automatically adjusting the smoothing factor based on transaction frequency. This ensures fair scoring across agents with different activity levels.
Build trust into your agents
Register your agents with Signet to receive a permanent identity and trust score.