Una-May O’Reilly

Genetic Programming: Theory and Practice

The talks and discussion during the fourth annual Genetic Programming Theory and Practice workshop (GPTPIV), held in Ann Arbor, Michigan, from May 11 to May 13 2006, suggest that the development of GP has crossed a watershed, from an emphasis on exploratory research to focusing on tackling large, real-world applications. Organized by the Center for the Study of Complex Systems (CSCS) of the University of Michigan and supported by Third Millenium, State Street Global Advisors (SSgA), Christopher T. May of Red Queen Capital Management, Michael Korn, and the Biocomputing and Developmental Systems Group of the University of Limerick, the goal of the workshop is to bridge the gap between theory and practice. Paraphrasing the introduction to the first workshop, the goal is “to allow theory to inform practice and practice to test theory.” To that end, the GPTP workshop again assembled a group of leading theoreticians and practitioners to present and discuss their recent work.

Genetic Programming Theory and Practice II

Genetic Programming Theory and Practice.- Discovering Financial Technical Trading Rules Using.- Abstraction GP.- Using Genetic Programming in Industrial Statistical Model Building Population Sizing for Genetic Programming.- Considering the Roles of Structure in Problem Solving by Computer.- Lessons Learned using Genetic Programming in a Stock Picking Context Favourable Biasing of Function Sets.- Toward Automated Design of Industrial-Strength Analog Circuits by Means of Genetic Programming.- Topological Synthesis of Robust Systems.- Does Genetic Programming Inherently Adopt Structured DesignTechniques?- Genetic Programming of an Algorithmic Chemistry.- ACGP: Adaptable Constrained Genetic Programming.- Searching for Supply Chain Reordering Policies.- Cartesian Genetic Programming and the Post Docking Filtering Problem.- Listening to Data: Tuning a Genetic Programming System.- Incident Detection on Highways.- Pareto-Front Exploitation in Symbolic Regression.- An Evolved Antenna for a NASA Mission.

An Interactive Artificial Ant Approach to Non-photorealistic Rendering

We couple artificial ant and computer graphics techniques to create an approach to Non-Photorealistic Rendering (NPR). A user interactively takes turns with an artificial ant colony to transform a photograph into a stylized picture. In turn with the user specifying its control parameters, members of a colony of artificial ants scan the source image locally, following strong edges or wandering around flat zones, and draw marks on the canvas depending on their discoveries. Among a variety of obtained effects, two are painterly rendering and pencil sketching.

Autotuning algorithmic choice for input sensitivity

A daunting challenge faced by program performance autotuning is input sensitivity, where the best autotuned configuration may vary with different input sets. This paper presents a novel two-level input learning algorithm to tackle the challenge for an important class of autotuning problems, algorithmic autotuning. The new approach uses a two-level input clustering method to automatically refine input grouping, feature selection, and classifier construction. Its design solves a series of open issues that are particularly essential to algorithmic autotuning, including the enormous optimization space, complex influence by deep input features, high cost in feature extraction, and variable accuracy of algorithmic choices. Experimental results show that the new solution yields up to a 3x speedup over using a single configuration for all inputs, and a 34x speedup over a traditional one-level method for addressing input sensitivity in program optimizations.

Genr8: Architects’ Experience with an Emergent Design Tool

We present the computational design tool Genr8 and six different architectural projects making extensive use of Genr8. Genr8 is based on ideas from Evolutionary Computation (EC) and Artificial Life and it produces surfaces using an organic growth algorithm inspired by how plants grow. These algorithms have been implemented as an architect’s design tool and the chapter provides an illustration of the possibilities that the tool provides.

GRACE: generative robust analog circuit exploration

We motivate and describe an analog evolvable hardware design platform named GRACE (i.e. Generative Robust Analog Circuit Exploration). GRACE combines coarse-grained, topological circuit search with intrinsic testing on a Commercial Off-The-Shelf (COTS) field programmable device, the Anadigm AN221E04. It is suited for adaptive, fault tolerant system design as well as CAD flow applications.

Evolutionary optimization of flavors

We have acquired panelist data that provides hedonic (liking) ratings for a set of 40 flavors each composed of the same 7 ingredients at different concentration levels. Our goal is to use this data and predict other flavors, composed of the same ingredients in new combinations, which the panelist will like. We describe how we first employ Pareto-Genetic Programming (GP) to generate a surrogate for the human panelist from the 40 observations. This surrogate, in fact an ensemble of GP symbolic regression models, can predict liking scores for flavors outside the observations and provide a confidence in the prediction. We then employ a multi-objective particle swarm optimization (MOPSO) to design a well and consistently liked flavor suite for a panelist. The MOPSO identifies flavors that are well liked, i.e., high liking score, and consistently-liked, i.e., of maximum confidence. Further, we generate flavors that are well and consistently liked by a cluster of panelists, by giving the MOPSO slightly different objectives.

Population Sizing for Genetic Programming Based on Decision-Making

This chapter derives a population sizing relationship for genetic programming (GP). Following the population-sizing derivation for genetic algorithms in (Goldberg et al., 1992), it considers building block decision-making as a key facet. The analysis yields a GP-unique relationship because it has to account for bloat and for the fact that GP solutions often use subsolutions multiple times. The population-sizing relationship depends upon tree size, solution complexity, problem difficulty and building block expression probability. The relationship is used to analyze and empirically investigate population sizing for three model GP problems named ORDER, ON-OFF and LOUD. These problems exhibit bloat to differing extents and differ in whether their solutions require the use of a building block multiple times.

EC-Star: A Massive-Scale, Hub and Spoke, Distributed Genetic Programming System

We describe a new Genetic Programming systemnamed EC-Star. It is supported by an open infrastructure, commercial-volunteer-client parallelization framework. The framework enables robust and massive-scale evolution and motivates the hub and spoke network topology of EC-Star’s distributed GP model. In this model an Evolution Coordinator occupies the hub and an Evolutionary Engine occupies each spoke. The Evolution Coordinator uses a layered framework to dispatch high performing, partially evaluated candidate solutions for additional fitness-case exposure, genetic mixing, and evolution to its Evolutionary Engines. It operates asynchronously with each Evolutionary Engine and never blocks waiting for results from an Evolutionary Engine.

Adapting Convergent Scheduling Using Machine-Learning

Convergent scheduling is a general framework for instruction scheduling and cluster assignment for parallel, clustered architectures. A convergent scheduler is composed of many independent passes, each of which implements a specific compiler heuristic. Each of the passes shares a common interface, which allows them to be run multiple times, and in any order. Because of this, a convergent scheduler is presented with a vast number of legal pass orderings. In this work, we use machine-learning techniques to automatically search for good orderings. We do so by evolving, through genetic programming, s-expressions that describe a particular pass sequence. Our system has the flexibility to create dynamic sequences where the ordering of the passes is predicated upon characteristics of the program being compiled. In particular, we implemented a few tests on the present state of the code being compiled. We are able to find improved sequences for a range of clustered architectures. These sequences were tested with cross-validation, and generally outperform Desoli’s PCC and UAS.