Search

Classification algorithms can suffer from performance degradation when the class distribution is unbalanced. This paper develops a two-step approach to evolving ensembles using genetic programming (GP) for unbalanced data. The first step uses multiobjective (MO) GP to evolve a Pareto-approximated front of GP classifiers to form the ensemble by trading-off the minority and the majority class against each other during learning. The MO component alleviates the reliance on sampling to artificially rebalance the data. The second step, which is the focus this paper, proposes a novel ensemble selection approach using GP to automatically find/choose the best individuals for the ensemble. This new GP approach combines multiple Pareto-approximated front members into a single composite genetic program solution to represent the (optimized) ensemble. This ensemble representation has two main advantages/novelties over traditional genetic algorithm (GA) approaches. First, by limiting the depth of the composite solution trees, we use selection pressure during evolution to find small highly-cooperative groups of individuals for the ensemble. This means that ensemble sizes are not fixed a priori (as in GA), but vary depending on the strength of the base learners. Second, we compare different function set operators in the composite solution trees to explore new ways to aggregate the member outputs and thus, control how the ensemble computes its output. We show that the proposed GP approach evolves smaller more diverse ensembles compared to an established ensemble selection algorithm, while still performing as well as, or better than the established approach. The evolved GP ensembles also perform well compared to other bagging and boosting approaches, particularly on tasks with high levels of class imbalance. [ABSTRACT FROM AUTHOR]

Morphological and attribute profiles have been proven to be effective tools to fuse spectral and spatial information for classification of remote sensing data. A wide range of filters (i.e., number of levels in the profiles) is usually necessary in order to properly model the spatial information in a remote sensing scene. A dense sampling of the values of the parameters of the filters generates profiles that have both a very large dimensionality (leading to the Hughes phenomenon in classification) and a high redundancy. In this paper, a novel iterative technique based on genetic algorithms (GAs) is proposed to automatically optimize the selection of the optimal features from the profiles. The selection of the filtered images that compose the profile is performed by dividing them into three classes corresponding to high, medium, and low importance. We propose to measure the importance (modeled in terms of discriminative power in the classification task) using a random forest classifier, which provides a rank for each feature with its model. Only the set of images associated with the highest importance is selected, i.e., preserved for classification. The proposed technique is applied to the features labeled with medium importance, whereas the images with the lowest importance are removed from the profile. This method is employed to classify three hyperspectral data sets achieving significantly high classification accuracy values. A parallel computing implementation has been developed in order to significantly reduce the time required for the run of the GAs. [ABSTRACT FROM AUTHOR]

Estimation of distribution algorithms are gaining increased research interest due to their advantage in exploiting linkage information. This paper examines the sampling techniques of a restricted Boltzmann machine-based multi-objective (MO) estimation of distribution algorithm (REDA). The behaviors of the sampling techniques in terms of energy levels are rigorously investigated, and a sampling mechanism that exploits the energy information of the solutions in a trained network is proposed to improve the search capability of the algorithm. The REDA is then hybridized, with a genetic algorithm and a local search based on an evolutionary gradient approach, to enhance the exploration and exploitation capabilities of the algorithm. Thirty-one benchmark test problems, which consist of different difficulties and characteristics, are used to examine the efficiency of the proposed algorithm. Empirical studies show that the proposed algorithm gives promising results in terms of inverted generational distance and nondominance ratio in most of the test problems. [ABSTRACT FROM AUTHOR]

Neural architecture search (NAS), which automatically designs the architectures of deep neural networks, has achieved breakthrough success over many applications in the past few years. Among different classes of NAS methods, evolutionary computation-based NAS (ENAS) methods have recently gained much attention. Unfortunately, the development of ENAS is hindered by unfair comparison between different ENAS algorithms due to different training conditions and high computational cost caused by expensive performance evaluation. This article develops a platform named BenchENAS, in short for benchmarking evolutionary NAS, to address these issues. BenchENAS makes it easy to achieve fair comparisons between different algorithms by keeping them under the same settings. To accelerate the performance evaluation in a common lab environment, BenchENAS designs a novel and generic efficient evaluation method for the population characteristics of evolutionary computation. This method has greatly improved the efficiency of the evaluation. Furthermore, BenchENAS is easy to install and highly configurable and modular, which brings benefits in good usability and easy extensibility. This article conducts efficient comparison experiments on eight ENAS algorithms with high GPU utilization on this platform. The experiments validate that the fair comparison issue does exist in the current ENAS algorithms, and BenchENAS can alleviate this issue. A Website has been built to promote BenchENAS at https://benchenas.com , where interested researchers can obtain the source code and document of BenchENAS for free. [ABSTRACT FROM AUTHOR]

Genetic programming (GP) has been applied to feature learning in image classification and achieved promising results. However, one major limitation of existing GP-based methods is the high computational cost, which may limit their applications on large-scale image classification tasks. To address this, this article develops a divide-and-conquer GP algorithm with knowledge transfer (KT) and ensembles to achieve fast feature learning in image classification. In the new algorithm framework, a divide-and-conquer strategy is employed to split the training data and the population into small subsets or groups to reduce computational time. A new KT method is proposed to improve GP learning performance. A new fitness function based on log loss and a new ensemble formulation strategy are developed to build an effective ensemble for image classification. The performance of the proposed approach has been examined on 12 image classification datasets of varying difficulty. The results show that the new approach achieves better classification performance in significantly less computation time than the baseline GP-based algorithm. The comparisons with state-of-the-art algorithms show that the new approach achieves better or comparable performance in almost all the comparisons. Further analysis demonstrates the effectiveness of ensemble formulation and KT in the proposed approach. [ABSTRACT FROM AUTHOR]