Waleed Yamany

Moth-flame optimization for training Multi-Layer Perceptrons

Multi-Layer Perceptron (MLP) is one of the Feed-Forward Neural Networks (FFNNs) types. Searching for weights and biases in MLP is important to achieve minimum training error. In this paper, Moth-Flame Optimizer (MFO) is used to train Multi-Layer Perceptron (MLP). MFO-MLP is used to search for the weights and biases of the MLP to achieve minimum error and high classification rate. Five standard classification datasets are utilized to evaluate the performance of the proposed method. Moreover, three function-approximation datasets are used to test the performance of the proposed method. The proposed method (i.e. MFO-MLP) is compared with four well-known optimization algorithms, namely, Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), and Evolution Strategy (ES). The experimental results prove that the MFO algorithm is very competitive, solves the local optima problem, and it achieves a high accuracy.

A New Multi-layer Perceptrons Trainer Based on Ant Lion Optimization Algorithm

In this paper, Ant Lion Optimizer (ALO) was presented to train Multi-Layer Perceptron (MLP). ALO was used to find the weights and biases of the MLP to achieve a minimum error and a high classification rate. Four standard classification datasets were used to benchmark the performance of the proposed method. In addition, the performance of the proposed method were compared with three well-known optimization algorithms, namely, Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Ant Colony Optimization (ACO). The experimental results showed that the ALO algorithm with the MLP was very competitive as it solved the local optima problem and achieved a high accuracy rate.

Hybrid flower pollination algorithm with rough sets for feature selection

Flower pollination algorithm (FPA) optimization is a new evolutionary computation technique that inspired from the pollination process of flowers. In this paper, a model for multi-objective feature selection based on flower pollination algorithm (FPA) optimization hybrid with rough set is proposed. The proposed model exploits the capabilities of filter-based feature selection and wrapper-based feature selection. Filter-based approach can be described as data oriented methods that not directly related to classification performance. Wrapper-based approach is more related to classification performance but it does not face redundancy and dependency among the selected feature set. Therefore, we proposed a multi-objective fitness function that uses FPA to the find optimal feature subset. The multi-objective fitness function enhances classification performance and guarantees minimum redundancy among selected features. At begin of the optimization process, fitness function uses mutual information among feature as a goal for optimization. While at some later time and using the same population, the fitness function is switched to be more classifier dependent and hence exploits rough-set classifier as a guide to classification performance. The proposed model was tested on eight datasets form UCI data repository and proves advance over other search methods as particle swarm optimization (PSO) and genetic algorithm (GA).

New approach for feature selection based on rough set and bat algorithm

This paper presents a new feature selection technique based on rough sets and bat algorithm (BA). BA is attractive for feature selection in that bats will discover best feature combinations as they fly within the feature subset space. Compared with GAs, BA does not need complex operators such as crossover and mutation, it requires only primitive and simple mathematical operators, and is computationally inexpensive in terms of both memory and runtime. A fitness function based on rough-sets is designed as a target for the optimization. The used fitness function incorporates both the classification accuracy and number of selected features and hence balances the classification performance and reduction size. This paper make use of four initialisation strategies for starting the optimization and studies its effect on bat performance. The used initialization reflects forward and backward feature selection and combination of both. Experimentation is carried out using UCI data sets which compares the proposed algorithm with a GA-based and PSO approaches for feature reduction based on rough-set algorithms. The results on different data sets shows that bat algorithm is efficient for rough set-based feature selection. The used rough-set based fitness function ensures better classification result keeping also minor feature size.

Wolf search algorithm for attribute reduction in classification

Data sets ordinarily includes a huge number of attributes, with irrelevant and redundant attributes. Redundant and irrelevant attributes might minimize the classification accuracy because of the huge search space. The main goal of attribute reduction is choose a subset of relevant attributes from a huge number of available attributes to obtain comparable or even better classification accuracy than using all attributes. A system for feature selection is proposed in this paper using a modified version of the wolf search algorithm optimization. WSA is a bio-inspired heuristic optimization algorithm that imitates the way wolves search for food and survive by avoiding their enemies. The WSA 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 is applied on a set of the UCI machine learning data sets and proves good performance in comparison with the GA and PSO optimizers commonly used in this context.

An Innovative Approach for Attribute Reduction Using Rough Sets and Flower Pollination Optimisation

Optimal search is a major challenge for wrapper-based attribute reduction. Rough sets have been used with much success, but current hill-climbing rough set approaches to attribute reduction are insufficient for finding optimal solutions. In this paper, we propose an innovative use of an intelligent optimisation method, namely the flower search algorithm (FSA), with rough sets for attribute reduction. FSA is a relatively recent computational intelligence algorithm, which is inspired by the pollination process of flowers. For many applications, the attribute space, besides being very large, is also rough with many different local minima which makes it difficult to converge towards an optimal solution. FSA can adaptively search the attribute space for optimal attribute combinations that maximise a given fitness function, with the fitness function used in our work being rough set-based classification. Experimental results on various benchmark datasets from the UCI repository confirm our technique to perform well in comparison with competing methods.

New Rough Set Attribute Reduction Algorithm Based on Grey Wolf Optimization

In this paper, we propose a new attribute reduction strategy based on rough sets and grey wolf optimization (GWO). Rough sets have been used as an attribute reduction technique with much success, but current hill-climbing rough set approaches to attribute reduction are inconvenient at finding optimal reductions as no perfect heuristic can guarantee optimality. Otherwise, complete searches are not feasible for even medium sized datasets. So, stochastic approaches provide a promising attribute reduction technique. Like Genetic Algorithms, GWO is a new evolutionary computation technique, mimics the leadership hierarchy and hunting mechanism of grey wolves in nature. The grey wolf optimization find optimal regions of the complex search space through the interaction of individuals in the population. Compared with GAs, GWO does not need complex operators such as crossover and mutation, it requires only primitive and easy mathematical operators, and is computationally inexpensive in terms of both memory and runtime. Experimentation is carried out, using UCI data, which compares the proposed algorithm with a GA-based approach and other deterministic rough set reduction algorithms. The results show that GWO is efficient for rough set-based attribute reduction.

Multi-Objective Cuckoo Search Optimization for Dimensionality Reduction

Commonly, attributes in data sets are originally correlated, noisy and redundant. Thus, attribute reduction is a challenging task as it substantially affects the overall classification accuracy. In this research, a system for attribute reduction was proposed using correlation-based filter model for attribute reduction. The cuckoo search (CS) optimization algorithm was utilized to search the attribute space with minimum correlation among selected attributes. Then, the initially selected solutions, guaranteed to have minor correlation, are candidates for further improvement towards the classification accuracy fitness function. The performance of the proposed system has been tested via implementing it using various data sets. Also, its performance have has been compared against other common attribute reduction algorithms. Experimental results showed that the proposed multi-objective CS system has outperformed the typical single-objective CS optimizer as well as outperforming both the particle swarm optimization (PSO) and genetic algorithm (GA) optimization algorithms.

Attribute reduction approach based on modified flower pollination algorithm

Attribute reduction approach is proposed in this paper based on a modified version of the flower pollination algorithm optimization (FPA). Flower pollination algorithm (FPA) is one of recently evolutionary computation technique, inspired by the pollination process of flowers. The modified FPA algorithm adaptively balance the exploration and exploitation to quickly find the optimal solution through using local searching with adaptive search diversity. The modified FPA 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 is applied on a eight dataset from the UCI machine learning data sets and proves a good performance in comparison with the genetic algorithm (GA) and particle swarm optimization (PSO), that commonly used in this context.

Rough Power Set Tree for Feature Selection and Classification: Case Study on MRI Brain Tumor

6 Scientific Research Group in Egypt (SRGE) http://www.egyptscience.net Abstract. This article presents a feature selection and classification sys- tem for 2D brain tumors from Magnetic resonance imaging (MRI) im- ages. The proposed feature selection and classification approach consists of four main phases. Firstly, clustering phase that applies the K-means clustering algorithm on 2D brain tumors slices. Secondly, feature ex- traction phase that extracts the optimum feature subset via using the brightness and circularity ratio. Thirdly, reduct generation phase that uses rough set based on power set tree algorithm to choose the reduct. Finally, classification phase that applies Multilayer Perceptron Neural Network algorithm on the reduct. Experimental results showed that the proposed classification approach achieved a high recognition rate com- pared to other classifiers including Naive Bayes, AD-tree and BF-tree.