Bayesian Adaptive Exploration

Abstract

I describe a framework for adaptive scientific exploration based on iterating an Observation--Inference--Design cycle that allows adjustment of hypotheses and observing protocols in response to the results of observation on-the-fly, as data are gathered. The framework uses a unified Bayesian methodology for the inference and design stages: Bayesian inference to quantify what we have learned from the available data and predict future data, and Bayesian decision theory to identify which new observations would teach us the most. When the goal of the experiment is simply to make inferences, the framework identifies a computationally efficient iterative ``maximum entropy sampling'' strategy as the optimal strategy in settings where the noise statistics are independent of signal properties. Results of applying the method to two ``toy'' problems with simulated data--measuring the orbit of an extrasolar planet, and locating a hidden one-dimensional object--show the approach can significantly improve observational efficiency in settings that have well-defined nonlinear models. I conclude with a list of open issues that must be addressed to make Bayesian adaptive exploration a practical and reliable tool for optimizing scientific exploration.

17 pages, 5 figures

Cited by in corpus (17)

• Cosmology and Fundamental Physics with the Euclid Satellite
• Simulation-based optimal Bayesian experimental design for nonlinear systems
• Robust Online Hamiltonian Learning
• Learning Quantum Systems
• Gradient-based stochastic optimization methods in Bayesian experimental design
• Forecasting the Bayes factor of a future observation
• Machine learning \& artificial intelligence in the quantum domain
• Parallel Algorithm for Solving Kepler's Equation on Graphics Processing Units: Application to Analysis of Doppler Exoplanet Searches
• Rotating Stars and Revolving Planets: Bayesian Exploration of the Pulsating Sky
• Optimal Experimental Design Using A Consistent Bayesian Approach
• Designing Intelligent Instruments
• Robust Bayesian Target Value Optimization
• Prospects for Refining Kepler TTV Masses using TESS Observations
• An adaptive scheduling tool to optimize measurements to reach a scientific objective: methodology and application to the measurements of stellar orbits in the Galactic Center
• Detecting Extra-solar Planets with a Bayesian hybrid MCMC Kepler periodogram
• Efficient adaptive Bayesian estimation of a slowly fluctuating Overhauser field gradient
• $\mathtt{KOBEsim}$: a Bayesian observing strategy algorithm for planet detection in radial velocity blind-search surveys