Author

L. van der Laan

3 papers indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Preprint Aug 2026

Bellman Calibration for Marginalized Importance Weighting in Offline Reinforcement Learning

The introduction of isotonic Bellman calibration, a one-dimensional, model-agnostic post-processing method that reduces residual occupancy-balance violations while preserving the ranking information in any initial occupancy-ratio estimate, and establishes finite-sample calibration guarantees and a KL oracle inequality...

L. van der Laan, Nathan Kallus · 0 citations

#machine learning Preprint Dec 2025

Soft Fitted Q-Iteration without Bellman Completeness: Occupancy Reweighting and Temperature Annealing

Under \(Q\)-function realizability and local regularity, it is established that soft control locally inherits the contraction of policy evaluation in a discounted-occupancy norm, and local contraction and finite-sample convergence with estimated ratios are established.

L. van der Laan, Nathan Kallus · 3 citations

#machine learning Preprint Dec 2025

Fitted Q-Evaluation without Bellman Completeness via Occupancy Weighting

Combining occupancy-weighted FQE with fitted occupancy-ratio evaluation gives an end-to-end guarantee governed by the complexities and direct approximation errors of the value-function and occupancy-ratio classes, removing the need for Bellman completeness.

L. van der Laan, Nathan Kallus · 0 citations