Automatic Derivation of Statistical Data Analysis Algorithms: Planetary Nebulae and Beyond

Publisher Website
Google Scholar
Abstract
AutoBayes is a fully automatic program synthesis system for the data analysis domain. Its input is a declarative problem description in form of a statistical model; its output is documented and optimized C/C++ code. The synthesis process relies on the combination of three key techniques. Bayesian networks are used as a compact internal representation mechanism which enables problem decompositions and guides the algorithm derivation. Program schemas are used as independently composable building blocks for the algorithm construction; they can encapsulate advanced algorithms and data structures. A symbolic‐algebraic system is used to find closed‐form solutions for problems and emerging subproblems. In this paper, we describe the application of AutoBayes to the analysis of planetary nebulae images taken by the Hubble Space Telescope. We explain the system architecture, and present in detail the automatic derivation of the scientists’ original analysis as well as a refined analysis using clustering models. This study demonstrates that AutoBayes is now mature enough so that it can be applied to realistic scientific data analysis tasks.

This publication has 0 references indexed in Scilit: