Efficient A-Learning for Dynamic Treatment Regimes: A Handout
نویسندگان
چکیده
This handout describes an locally efficient version of an off-line advantage learning (ALearning) method for estimating dynamic treatment regimes using batch data based on (Murphy (2003); Robins (2004); Blatt et al. (2004)). The local efficient version is derived by Robins (Robins (2004)). These notes are meant to compliment the powerpoint slides found at http://www.stat.lsa.umich.edu/∼samurphy/seminars/. A-Learning can be seen as an alternative to Q-Learning (Watkins (1989); Watkins and Dayan (1992)). At this time, it is not known which method is superior for developing dynamic treatment regimes. For more details, including initial discussions concerning the relative merits of A-Learning versus Q-Learning, please refer to the slides and the above-cited papers.
منابع مشابه
A Review of Some Inferential Challenges in Constructing Dynamic Treatment Regimes
This handout reviews non-regularity problems that arise in the estimation of dynamic treatment regimes. The non-regularity makes it difficult to conduct inference on the estimated decision rules. When data sets that are small relative to the noise level, it is crucial to provide inferential methods and measures of confidence when estimating the decision rules. This handout illustrates how both ...
متن کاملA Bayesian Machine Learning Approach for Optimizing Dynamic Treatment Regimes
Medical therapy often consists of multiple stages, with a treatment chosen by the physician at each stage based on the patient’s history of treatments and clinical outcomes. These decisions can be formalized as a dynamic treatment regime. This paper describes a new approach for optimizing dynamic treatment regimes that bridges the gap between Bayesian inference and existing approaches, like Q-l...
متن کاملDeveloping a dynamic yield and growth model for saffron under different irrigation regimes
Better irrigation management and more efficient management of crop production require modeling of plant growth and crop yield. More applicable models are usually simple and requires less and accessible inputs. The objective of this study was to develop a model for growth and yield prediction of saffron under various irrigation regimes. In this modeling soil water budget and other simple rel...
متن کاملDynamic regime marginal structural mean models for estimation of optimal dynamic treatment regimes, Part I: main content.
Dynamic treatment regimes are set rules for sequential decision making based on patient covariate history. Observational studies are well suited for the investigation of the effects of dynamic treatment regimes because of the variability in treatment decisions found in them. This variability exists because different physicians make different decisions in the face of similar patient histories. I...
متن کاملDynamic Treatment Regimes.
A dynamic treatment regime consists of a sequence of decision rules, one per stage of intervention, that dictate how to individualize treatments to patients based on evolving treatment and covariate history. These regimes are particularly useful for managing chronic disorders, and fit well into the larger paradigm of personalized medicine. They provide one way to operationalize a clinical decis...
متن کامل