Fitness Causes Bloat

Langdon, William B.; Poli, Roberto

doi:10.1007/978-1-4471-0427-8_2

Cited by 133 publications

(118 citation statements)

References 6 publications

Supporting

Mentioning

115

Contrasting

Unclassified

Order By: Relevance

“…Finally, it is important to point out that even though operator equalisation can limit bloat for some problems, it is a computationally expensive algorithm, so developing other methods remains a worthwhile endeavor. This work revisits another bloat theory, which can be called the fitness-causesbloat theory (FCBT), developed by Langdon and Poli [13,14]. The main arguments of FCBT proceed as follows [13,14,32].…”

Section: Recent Advances In Bloatmentioning

confidence: 99%

“…This work revisits another bloat theory, which can be called the fitness-causesbloat theory (FCBT), developed by Langdon and Poli [13,14]. The main arguments of FCBT proceed as follows [13,14,32]. For a variable length GP representation (such as a tree based GP) many genotypically different programs, of different sizes, can produce the same outputs on a given set of fitness cases.…”

Section: Recent Advances In Bloatmentioning

confidence: 99%

“…This work revisits the fitness-causes-bloat theory of Langdon and Poli [13,14,31], that basically states that the search for better fitness will bias the search towards larger trees, simply because there are more large programs than small ones. Silva and Costa [31] state it clearly:…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Preliminary Study of Bloat in Genetic Programming with Behavior-Based Search

Trujillo

Naredo

Martínez

2013

EVOLVE - A Bridge Between Probability, Set Oriented Numerics, and Evolutionary Computation IV

View full text Add to dashboard Cite

Abstract. Bloat is one of the most interesting theoretical problems in genetic programming (GP), and one of the most important pragmatic limitations in the development of real-world GP solutions. Over the years, many theories regarding the causes of bloat have been proposed and a variety of bloat control methods have been developed. It seems that one of the underlying causes of bloat is the search for fitness; as the fitness-causes-bloat theory states, selective bias towards fitness seems to unavoidably lead the search towards programs with a large size. Intuitively, however, abandoning fitness does not appear to be an option. This paper, studies a GP system that does not require an explicit fitness function, instead it relies on behavior-based search, where programs are described by the behavior they exhibit and selective pressure is biased towards unique behaviors using the novelty search algorithm. Initial results are encouraging, the average program size of the evolving population does not increase with novelty search; i.e., bloat is avoided by focusing on novelty instead of quality.

show abstract

Section: Recent Advances In Bloatmentioning

confidence: 99%

Section: Recent Advances In Bloatmentioning

confidence: 99%

See 1 more Smart Citation

Preliminary Study of Bloat in Genetic Programming with Behavior-Based Search

Trujillo

Naredo

Martínez

2013

EVOLVE - A Bridge Between Probability, Set Oriented Numerics, and Evolutionary Computation IV

View full text Add to dashboard Cite

show abstract

“…On average the inserted subtree is bigger than the excised one, so such offspring are bigger than average while retaining the fitness of their parent, leading ultimately to growth in the average program size. Another important theory, the nature of program search spaces theory [17,19], predicts that above a certain size, the distribution of fitnesses does not vary with size. Since there are more long programs, the number of long programs of a given fitness is greater than the number of short programs of the same fitness.…”

Section: Bloat In Theorymentioning

confidence: 99%

Parsimony Pressure Made Easy: Solving the Problem of Bloat in GP

Poli

McPhee

2013

Theory and Principled Methods for the Design of Metaheuristics

View full text Add to dashboard Cite

The parsimony pressure method is perhaps the simplest and most frequently used method to control bloat in genetic programming. In this chapter we first reconsider the size evolution equation for genetic programming developed in [28] and rewrite it in a form that shows its direct relationship to Price's theorem. We then use this new formulation to derive theoretical results that show how to practically and optimally set the parsimony coefficient dynamically during a run so as to achieve complete control over the growth of the programs in a population. Experimental results confirm the effectiveness of the method, as we are able to tightly control the average program size under a variety of conditions. These include such unusual cases as dynamically varying target sizes so that the mean program size is allowed to grow during some phases of a run, while being forced to shrink in others.

show abstract

“…Poli's work on a GP schema theory (and later Langdon and Poli) [15][16][17][18] presumes schemata that are structurally based (i.e., has a certain size and depth). Bloating has also been associated with structure; it has been Langdon and Poli's contention that bloating occurs partly as a result of a random walk on a landscape of tree-shape distributions (e.g., [19,20]). …”

Section: Introductionmentioning

confidence: 99%

Visualizing Tree Structures in Genetic Programming

Daida

Hilss

Ward

et al. 2003

Genetic and Evolutionary Computation — GECCO 2003

View full text Add to dashboard Cite

Abstract. This paper presents methods to visualize the structure of trees that occur in genetic programming. These methods allow for the inspection of structure of entire trees of arbitrary size. The methods also scale to allow for the inspection of structure for an entire population. Examples are given from a typical problem. The examples indicate further studies that might be enabled by visualizing structure at these scales.

show abstract

Fitness Causes Bloat

Cited by 133 publications

References 6 publications

Preliminary Study of Bloat in Genetic Programming with Behavior-Based Search

Preliminary Study of Bloat in Genetic Programming with Behavior-Based Search

Parsimony Pressure Made Easy: Solving the Problem of Bloat in GP

Visualizing Tree Structures in Genetic Programming

Contact Info

Product

Resources

About