Academic
Publications
Advanced Genetic Programming Based Machine Learning

Advanced Genetic Programming Based Machine Learning,10.1007/s10852-007-9065-6,Journal of Mathematical Modelling and Algorithms,Stephan M. Winkler,Mich

Advanced Genetic Programming Based Machine Learning   (Citations: 14)
BibTex | RIS | RefWorks Download
A Genetic Programming based approach for solving classification problems is presented in this paper. Classification is understood as the act of placing an object into a set of categories, based on the object’s properties; classification algorithms are designed to learn a function which maps a vector of object features into one of several classes. This is done by analyzing a set of input-output examples (“training samples”) of the function. Here we present a method based on the theory of Genetic Algorithms and Genetic Programming that interprets classification problems as optimization problems: Each presented instance of the classification problem is interpreted as an instance of an optimization problem, and a solution is found by a heuristic optimization algorithm. The major new aspects presented in this paper are advanced algorithmic concepts as well as suitable genetic operators for this problem class (mainly the creation of new hypotheses by merging already existing ones and their detailed evaluation). The experimental part of the paper documents the results produced using new hybrid variants of Genetic Algorithms as well as investigated parameter settings. Graphical analysis is done using a novel multiclass classifier analysis concept based on the theory of Receiver Operating Characteristic curves.
Journal: Journal of Mathematical Modelling and Algorithms - JMMA , vol. 6, no. 3, pp. 455-480, 2007
Cumulative Annual
View Publication
The following links allow you to view full publications. These links are maintained by other sources not affiliated with Microsoft Academic Search.
    • ...Genetic Programming (GP) is an evolutionary technique which has been successful in building reliable and accurate classifiers to solve a range of classification problems [2][3][4]...
    • ...A typical fitness measure in classification is the success or error rate of a solution on the training examples [3][14][16]...
    • ...These new objectives measure the “level of error” for each class, and are designed to add a finer-grain to the fitness landscape – this can guide greedy hillclimbing search better [3]...

    Urvesh Bhowanet al. Genetic Programming for Classification with Unbalanced Data

    • ...Genetic Programming (GP) is an evolutionary ML technique which has been successful in evolving reliable and accurate classifiers [4][5][6]...
    • ...However, GP, like many other ML approaches, can evolve “biased” classifiers, that is, solutions with strong majority class accuracy but poor minority class accuracy, when data sets are unbalanced [2][5][6]...
    • ...approximating the area under the Receiver Operating Characteristics (ROC) curve (known as the AUC) in fitness [5][6], or using fixed misclassification costs for class examples to boost classification rates [2][7]...
    • ...In binary classification, GP classifiers are usually represented as mathematical expressions, where the numeric program output is mapped to two class labels using a fixed class-threshold (zero) [5][7]...
    • ...to the majority class if the classifier output is negative, otherwise it will be assigned to the minority class [5][7]...
    • ...Fitness Function 2: Research has shown that fitness function Std can favour the evolution of classifiers biased toward the majority class [5][6][7]...

    Urvesh Bhowanet al. A Comparison of Classification Strategies in Genetic Programming with ...

    • ...In contrast to this, the possibility of self-adaptive selection pressure steering within SASEGASA caused by offspring selection (OS) has proven extremely powerful especially for GP problem representations; this has already been shown in [52], [53], [55] and [54], and will also be treated in the discussion of experimental results of the present article (Sect...

    Stephan M. Winkleret al. Using Enhanced Genetic Programming Techniques for Evolving Classifiers...

    • ...Genetic Programming (GP) is a promising machine learning and search technique which has been successful in building reliable classifiers to solve classification problems [8] [9] [10]...
    • ...To counter the class imbalance in two benchmark binary classification problems, Winker et al. [10] used the weighted sum of three distinct performance criteria as a fitness measure: the root mean squared error (RMS) in training (weighted the highest), a measure of the level of agreement between predicted and expected classifier values using genetic program ranges, and a separability measure similar to the AUC...
    • ...The idea here is that the more concise the classification model for program p is, the better its fitness [10]...

    Urvesh Bhowanet al. Differentiating between individual class performance in Genetic Progra...

    • ...Genetic Programming (GP) is a machine learning and search technique which has been widely successful in solving various classification problems [1][2][3]...

    Urvesh Bhowanet al. Genetic programming for image classification with unbalanced data

Sort by: