PseudoInputPolicy¶
- class itergp.methods.policies.PseudoInputPolicy(pseudo_inputs)¶
Bases:
LinearSolverPolicyPseudo-input policy.
- Parameters
pseudo_inputs (backend.Array) – Pseudo inputs, also known as inducing points.
- Return type
None
Attributes Summary
Pseudo-inputs defining virtual observation locations.
Methods Summary
__call__(solver_state[, rng])Return an action for a given solver state.
Attributes Documentation
- pseudo_inputs¶
Pseudo-inputs defining virtual observation locations.
Methods Documentation
- __call__(solver_state, rng=None)[source]¶
Return an action for a given solver state.
- Parameters
solver_state (probnum.linalg.solvers.LinearSolverState) – Current state of the linear solver.
rng (Optional[SeedSequence]) – Random number generator.
- Returns
Next action to take.
- Return type
action