Loading...
Loading...
Ω
The component of a learning cycle defining which signals are visible or admissible as evidence. Narrow observation spaces filter out inconvenient evidence; politically filtered observation spaces privilege compliance metrics over performance data. Expanding Ω is necessary but insufficient for learning—update space must also be authorised.
A healthcare system with Ω = cost metrics only cannot learn about clinical outcomes. Expanding Ω to include patient outcomes enables learning only if the update space U also permits protocol changes.
ILA-Core Section 4