ABSTRACT
Given n variables to model, symbolic regression (SR) returns a flat list of n equations. As the number of state variables to be modeled scales, it becomes increasingly difficult to interpret such a list. Here we present a symbolic regression method that detects and captures hidden hierarchy in a given system. The method returns the equations in a hierarchical dependency graph, which increases the interpretability of the results. We demonstrate two variations of this hierarchical modeling approach, and show that both consistently outperform non-hierarchical symbolic regression on a number of synthetic data sets.
- S. V. Buldyrev, N. V. Dokholyan, S. Erramilli, M. Hong, J. Y. Kim, G. Malescio, and H. E. Stanley. Hierarchy in social organization. Physica A: Statistical Mechanics and its Applications, 330(3-4):653--659, 2003.Google ScholarCross Ref
- Hironori Hirata and Robert E. Ulanowicz. Information theoretical analysis of the aggregation and hierarchical structure of ecological networks. Journal of Theoretical Biology, 116(3):321--341, 1985.Google ScholarCross Ref
- Erzsébet Ravasz and Albert-László Barabási. Hierarchical organization in complex networks. Phys. Rev. E, 67:026112, Feb 2003.Google ScholarCross Ref
- Ralf J Sommer. Homology and the hierarchy of biological systems. Bioessays, 30(7):653--8, 2008.Google ScholarCross Ref
Index Terms
- Automatic identification of hierarchy in multivariate data
Recommendations
Hierarchical Means Clustering
AbstractIn the cluster analysis literature, there are several partitioning (non-hierarchical) methods for clustering multivariate objects based on model estimation. Distinct to these methods is the use of a system of n nested statistical models and the ...
Symbolic Regression Based Extreme Learning Machine Models for System Identification
AbstractReproducible machine learning models with less number of parameters and fast optimization are preferred in embedded system design for the applications of artificial intelligence. Due to implementation advantages, symbolic regression with genetic ...
Least squares parameter estimation and multi-innovation least squares methods for linear fitting problems from noisy data
AbstractLeast squares is an important method for solving linear fitting problems and quadratic optimization problems. This paper explores the properties of the least squares methods and the multi-innovation least squares methods. It ...
Comments