Eid Emary

Binary grey wolf optimization approaches for feature selection

In this work, a novel binary version of the grey wolf optimization (GWO) is proposed and used to select optimal feature subset for classification purposes. Grey wolf optimizer (GWO) is one of the latest bio-inspired optimization techniques, which simulate the hunting process of grey wolves in nature. The binary version introduced here is performed using two different approaches. In the first approach, individual steps toward the first three best solutions are binarized and then stochastic crossover is performed among the three basic moves to find the updated binary grey wolf position. In the second approach, sigmoidal function is used to squash the continuous updated position, then stochastically threshold these values to find the updated binary grey wolf position. The two approach for binary grey wolf optimization (bGWO) are hired in the feature selection domain for finding feature subset maximizing the classification accuracy while minimizing the number of selected features. The proposed binary versions were compared to two of the common optimizers used in this domain namely particle swarm optimizer and genetic algorithms. A set of assessment indicators are used to evaluate and compared the different methods over 18 different datasets from the UCI repository. Results prove the capability of the proposed binary version of grey wolf optimization (bGWO) to search the feature space for optimal feature combinations regardless of the initialization and the used stochastic operators.

Feature Subset Selection Approach by Gray-Wolf Optimization

Feature selection algorithm explores the data to eliminate noisy, irrelevant, redundant data, and simultaneously optimize the classification performance. In this paper, a classification accuracy-based fitness function is proposed by gray-wolf optimizer to find optimal feature subset. Gray-wolf optimizer is a new evolutionary computation technique which mimics the leadership hierarchy and hunting mechanism of gray wolves in nature. The aim of the gray wolf optimization is find optimal regions of the complex search space through the interaction of individuals in the population. Compared with particle swarm optimization (PSP) and Genetic Algorithms (GA) over a set of UCI machine learning data repository, the proposed approach proves better performance in both classification accuracy and feature size reduction. Moreover, the gray wolf optimization approach proves much robustness against initialization in comparison with PSO and GA optimizers.

Binary ant lion approaches for feature selection

In this paper, binary variants of the ant lion optimizer (ALO) are proposed and used to select the optimal feature subset for classification purposes in wrapper-mode. ALO is one of the recently bio-inspired optimization techniques that imitates the hunting process of ant lions. Moreover, ALO balances exploration and exploitation using a single operator that can adaptively searches the domain of solutions for the optimal solution. Binary variants introduced here are performed using two different approaches. The first approach takes only the inspiration of ALO operators and makes the corresponding binary operators. In the second approach, the native ALO is applied while its continuous steps are threshold using suitable threshold function after squashing them. The proposed approaches for binary ant lion optimizer (BALO) are utilized in the feature selection domain for finding feature subset that maximizing the classification performance while minimizing the number of selected features. The proposed binary algorithms were compared to three common optimization algorithms hired in this domain namely particle swarm optimizer (PSO), genetic algorithms (GAs), binary bat algorithm (BBA), as well as the native ALO. A set of assessment indicators is used to evaluate and compare the different methods over 21 data sets from the UCI repository. Results prove the capability of the proposed binary algorithms to search the feature space for optimal feature combinations regardless of the initialization and the used stochastic operators.

Feature Selection via Chaotic Antlion Optimization

Background Selecting a subset of relevant properties from a large set of features that describe a dataset is a challenging machine learning task. In biology, for instance, the advances in the available technologies enable the generation of a very large number of biomarkers that describe the data. Choosing the more informative markers along with performing a high-accuracy classification over the data can be a daunting task, particularly if the data are high dimensional. An often adopted approach is to formulate the feature selection problem as a biobjective optimization problem, with the aim of maximizing the performance of the data analysis model (the quality of the data training fitting) while minimizing the number of features used. Results We propose an optimization approach for the feature selection problem that considers a “chaotic” version of the antlion optimizer method, a nature-inspired algorithm that mimics the hunting mechanism of antlions in nature. The balance between exploration of the search space and exploitation of the best solutions is a challenge in multi-objective optimization. The exploration/exploitation rate is controlled by the parameter I that limits the random walk range of the ants/prey. This variable is increased iteratively in a quasi-linear manner to decrease the exploration rate as the optimization progresses. The quasi-linear decrease in the variable I may lead to immature convergence in some cases and trapping in local minima in other cases. The chaotic system proposed here attempts to improve the tradeoff between exploration and exploitation. The methodology is evaluated using different chaotic maps on a number of feature selection datasets. To ensure generality, we used ten biological datasets, but we also used other types of data from various sources. The results are compared with the particle swarm optimizer and with genetic algorithm variants for feature selection using a set of quality metrics.

Retinal blood vessel localization approach based on bee colony swarm optimization, fuzzy c-means and pattern search

The proposed vessel segmentation based on bee colony swarm optimization and fuzzy c-means is comprised of the following three fundamental building phases: Pre-processing phase: In the first phase of the investigation, was adapted to enhance the main brightness of the retina images before the actual segmentation is performed. Segmentation phase: In the second phase, bee colony swarm optimization with fuzzy cluster compactness fitness is applied to find clusters and further the obtained clusters are refined using pattern search with thinness fitness function. Post-processing phase: is applied for improving the segmentation accuracy by removing the non-thin connected components and gap filling. These three phases are described in detail in the following section along with the steps involved and the characteristics feature for each phase and the overall architecture of the introduced approach is described in the following figure.Display Omitted An automated retinal blood vessels segmentation approach based on two levels optimization principles is proposed.Uses the artificial bee colony optimization in conjunction with fuzzy cluster compactness fitness function.Pattern search is further used to enhance the segmentation results using shape description as a complementary feature. Accurate segmentation of retinal blood vessels is an important task in computer aided diagnosis and surgery planning of retinopathy. Despite the high resolution of photographs in fundus photography, the contrast between the blood vessels and retinal background tends to be poor. Furthermore, pathological changes of the retinal vessel tree can be observed in a variety of diseases such as diabetes and glaucoma. Vessels with small diameters are much liable to effects of diseases and imaging problems. In this paper, an automated retinal blood vessels segmentation approach based on two levels optimization principles is proposed. The proposed approach makes use of the artificial bee colony optimization in conjunction with fuzzy cluster compactness fitness function with partial belongness in the first level to find coarse vessels. The dependency on the vessel reflectance is problematic as the confusion with background and vessel distortions especially for thin vessels, so we made use of a second level of optimization. In the second level of optimization, pattern search is further used to enhance the segmentation results using shape description as a complementary feature. Thinness ratio is used as a fitness function for the pattern search optimization. The pattern search is a powerful tool for local search while artificial bee colony is a global search with high convergence speed. The proposed retinal blood vessels segmentation approach is tested on two publicly available databases DRIVE and STARE of retinal images. The results demonstrate that the performance of the proposed approach is comparable with state of the art techniques in terms of sensitivity, specificity and accuracy.

Retinal blood vessel segmentation using bee colony optimisation and pattern search

Accurate segmentation of retinal blood vessels is an important task in computer aided diagnosis of retinopathy. In this paper, we propose an automated retinal blood vessel segmentation approach based on artificial bee colony optimisation in conjunction with fuzzy c-means clustering. Artificial bee colony optimisation is applied as a global search method to find cluster centers of the fuzzy c-means objective function. Vessels with small diameters appear distorted and hence cannot be correctly segmented at the first segmentation level due to confusion with nearby pixels. We employ a pattern search approach to optimisation in order to localise small vessels with a different fitness function. The proposed algorithm is tested on the publicly available DRIVE and STARE retinal image databases and confirmed to deliver performance that is comparable with state-of-the-art techniques in terms of accuracy, sensitivity and specificity.

Impact of Chaos Functions on Modern Swarm Optimizers.

Exploration and exploitation are two essential components for any optimization algorithm. Much exploration leads to oscillation and premature convergence while too much exploitation slows down the optimization algorithm and the optimizer may be stuck in local minima. Therefore, balancing the rates of exploration and exploitation at the optimization lifetime is a challenge. This study evaluates the impact of using chaos-based control of exploration/exploitation rates against using the systematic native control. Three modern algorithms were used in the study namely grey wolf optimizer (GWO), antlion optimizer (ALO) and moth-flame optimizer (MFO) in the domain of machine learning for feature selection. Results on a set of standard machine learning data using a set of assessment indicators prove advance in optimization algorithm performance when using variational repeated periods of declined exploration rates over using systematically decreased exploration rates.

Firefly Optimization Algorithm for Feature Selection

In this paper, a system for feature selection based on firefly algorithm (FFA) optimization is proposed. Data sets ordinarily includes a huge number of attributes, with irrelevant and redundant attributes. Redundant and irrelevant attributes might reduce the classification accuracy because of the large search space. The main goal of attribute reduction is to choose a subset of relevant attributes from a huge number of available attributes to obtain comparable or even better classification accuracy from using all attributes. A system for feature selection is proposed in this paper using a modified version of the firefly algorithm (FFA) optimization. The modified FFA algorithm adaptively balance the exploration and exploitation to quickly find the optimal solution. FFA is a new evolutionary computation technique, inspired by the flash lighting process of fireflies. The FFA can quickly search the feature space for optimal or near-optimal feature subset minimizing a given fitness function. The proposed fitness function used incorporate both classification accuracy and feature reduction size. The proposed system was tested on eighteen data sets and proves advance over other search methods as particle swarm optimization (PSO) and genetic algorithm (GA) optimizers commonly used in this context using different evaluation indicators.

Retinal vessel segmentation based on possibilistic fuzzy c-means clustering optimised with cuckoo search

Automated analysis of retinal vessels is essential for the diagnosis of a wide range of eye diseases and plays an important role in automatic retinal disease screening systems. In this paper, we present an approach to automatic vessel segmentation in retinal images that utilises possibilistic fuzzy c-means (PFCM) clustering to overcome the problems of the conventional fuzzy c-means objective function. In order to obtain optimised clustering results using PFCM, a cuckoo search method is used. The cuckoo search algorithm, which is based on the brood parasitic behaviour of some cuckoo species in combination with the Levy flight behaviour of some birds and fruit flies, is applied to drive the optimisation of the fuzzy clustering. The performance of our algorithm is analysed on two benchmark databases, the DRIVE and STARE datasets, and encouraging segmentation performance is observed.

Experienced Gray Wolf Optimization Through Reinforcement Learning and Neural Networks

In this paper, a variant of gray wolf optimization (GWO) that uses reinforcement learning principles combined with neural networks to enhance the performance is proposed. The aim is to overcome, by reinforced learning, the common challenge of setting the right parameters for the algorithm. In GWO, a single parameter is used to control the exploration/exploitation rate, which influences the performance of the algorithm. Rather than using a global way to change this parameter for all the agents, we use reinforcement learning to set it on an individual basis. The adaptation of the exploration rate for each agent depends on the agent’s own experience and the current terrain of the search space. In order to achieve this, experience repository is built based on the neural network to map a set of agents’ states to a set of corresponding actions that specifically influence the exploration rate. The experience repository is updated by all the search agents to reflect experience and to enhance the future actions continuously. The resulted algorithm is called experienced GWO (EGWO) and its performance is assessed on solving feature selection problems and on finding optimal weights for neural networks algorithm. We use a set of performance indicators to evaluate the efficiency of the method. Results over various data sets demonstrate an advance of the EGWO over the original GWO and over other metaheuristics, such as genetic algorithms and particle swarm optimization.