Hierarquical structure over BSR missing block variables
User should be able to define latent, prior and generic constraints over missing block variables.
A tipical situation is that an input or output is non stationary, so scalar truncated normal distributions are no too apropriate.
Then, it should be possible to define an ARIMA model for the residuals of a special node missingResidual with linear equations of regression:
knownValue[n] = missingResidual[n] //If n-th datum is known
0 = missingResidual[n] - missingVariable[n] //If n-th datum is unknown
Change History (8)
| Component: |
Database →
Math
|
| Owner: |
changed from Jorge to Víctor de Buen Remiro
|
| Priority: |
highest →
normal
|
| Severity: |
blocker →
normal
|
| Type: |
defect →
task
|
| Milestone: |
→ BSR hierarchy and priors
|
| Version: |
→ 2.0.1
|
| Priority: |
normal →
highest
|
| Severity: |
normal →
blocker
|
| Description: |
modified (diff)
|
| Status: |
new →
accepted
|
| Description: |
modified (diff)
|
| Description: |
modified (diff)
|
| Description: |
modified (diff)
|
| Resolution: |
→ duplicate
|
| Status: |
accepted →
closed
|
A best solution is suplyied in ticket #759