Combining Analytical and Evolutionary Inductive Programming
نویسندگان
چکیده
Analytical inductive programming and evolutionary inductive programming are two opposing strategies for learning recursive programs from incomplete specifications such as input/output examples. Analytical inductive programming is data-driven, namely, the minimal recursive generalization over the positive input/output examples is generated by recurrence detection. Evolutionary inductive programming, on the other hand, is based on searching through hypothesis space for a (recursive) program which performs sufficiently well on the given input/output examples with respect to some measure of fitness. While analytical approaches are fast and guarantee some characteristics of the induced program by construction (such as minimality and termination) the class of inducable programs is restricted to problems which can be specified by few positive examples. The scope of programs which can be generated by evolutionary approaches is, in principle, unrestricted, but generation times are typically high and there is no guarantee that such a program is found for which the fitness is optimal. We present a first study exploring possible benefits from combining analytical and evolutionary inductive programming. We use the analytical system Igor2 to generate skeleton programs which are used as initial hypotheses for the evolutionary system Adate. We can show that providing such constraints can reduce the induction time of Adate.
منابع مشابه
Evolutionary Programming Guided by Analytically Generated Seeds
Evolutionary programming is the most powerful method for inducing recursive functional programs from input/output examples while taking into account efficiency and complexity constraints for the target program. However, synthesis time can be considerably high. A strategy which is complementary to the generate-and -test based approaches of evolutionary programming is inductive analytical program...
متن کاملExpansion of the GeLog System by Automatic Adaptation of Parameters
Different techniques have been introduced for the improvement of Evolutionary Algorithms (EA) to increase their performance as well as to achieve better results. Adaptive methods are commonly used. In the scope of this work these extensions are applied to GeLog, a system which implements a combination of two different approaches for automatic programming: inductive logic programming and genetic...
متن کاملGenetic Programming and Deductive-Inductive Learning: A Multi-Strategy Approach
Genetic Programming (GP) is a machine learning technique that was not conceived to use domain knowledge for generating new candidate solutions. It has been shown that GP can beneet from domain knowledge obtained by other machine learning methods with more powerful heuristics. However, it is not obvious that a combination of GP and a knowledge intensive machine learning method can work better th...
متن کاملSimultaneous Inductive and Deductive Modeling of Ecological Systems via Evolutionary Computation and Information Theory
SIMUL ©2006 The addition of two emerging technologies (evolutionary computation and ecoinformatics) to computational ecology can advance our ability to build better ecological models and thus deepen our understanding of the mechanistic complexity of ecological systems. This article describes one feasible approach toward this goal—the combining of inductive and deductive modeling techniques with...
متن کاملInductive Rule Learning on the Knowledge
We present an approach to learning sets of recursive rules based on analytical inductive programming. We propose that our approach can be used within cognitive architectures to model regularity detection and generalization over experience. Induced rule sets represent the knowledge underlying systematic behavior in complex situations. Such rule sets can explain systematicity and productivity of ...
متن کامل