An improved gravitational search algorithm for dynamic neural network identification

Gravitational search algorithm (GSA) is a newly developed and promising algorithm based on the law of gravity and interaction between masses. This paper proposes an improved gravitational search algorithm (IGSA) to improve the performance of the GSA, and first applies it to the field of dynamic neural network identification. The IGSA uses trial-and-error method to update the optimal agent during the whole search process. And in the late period of the search, it changes the orbit of the poor agent and searches the optimal agent’s position further using the coordinate descent method. For the experimental verification of the proposed algorithm, both GSA and IGSA are testified on a suite of four well-known benchmark functions and their complexities are compared. It is shown that IGSA has much better efficiency, optimization precision, convergence rate and robustness than GSA. Thereafter, the IGSA is applied to the nonlinear autoregressive exogenous (NARX) recurrent neural network identification for a magnetic levitation system. Compared with the system identification based on gravitational search algorithm neural network (GSANN) and other conventional methods like BPNN and GANN, the proposed algorithm shows the best performance.     

基于引力搜索机制的数据聚类及特征选择算法

为同步选择具有相关特征的数据聚类数量,提出一种基于引力搜索机制的聚类和特征选择算法.设计一种代理表示策略实现聚类中心和特征数量的编码;提出一种动态临界值方法决定聚类和特征数量,通过代理适应度的不断评估寻找最优聚类量和相关特征;分析算法的时间复杂度,通过8个经典数据集测试算法性能,并与7种常规数据聚类算法作对比.实验结果...     

Automatic clustering and feature selection using gravitational search algorithm and its application to microarray data analysis

Neural Computing and Applications - This paper proposes a novel approach that selects the number of clusters along with relevant features automatically and simultaneously. Gravitational search...     

Hybrid feature selection algorithm using symmetrical uncertainty and a harmony search algorithm

Microarray technology can be used as an efficient diagnostic system to recognise diseases such as tumours or to discriminate between different types of cancers in normal tissues. This technology has received increasing attention from the bioinformatics community because of its potential in designing powerful decision-making tools for cancer diagnosis. However, the presence of thousands or tens of thousands of genes affects the predictive accuracy of this technology from the perspective of classification. Thus, a key issue in microarray data is identifying or selecting the smallest possible set of genes from the input data that can achieve good predictive accuracy for classification. In this work, we propose a two-stage selection algorithm for gene selection problems in microarray data-sets called the symmetrical uncertainty filter and harmony search algorithm wrapper (SU-HSA). Experimental results show that the SU-HSA is better than HSA in isolation for all data-sets in terms of the accuracy and achieves a lower number of genes on 6 out of 10 instances. Furthermore, the comparison with state-of-the-art methods shows that our proposed approach is able to obtain 5 (out of 10) new best results in terms of the number of selected genes and competitive results in terms of the classification accuracy.     

An improved binary sparrow search algorithm for feature selection in data classification

Neural Computing and Applications - Feature Selection (FS) is an important preprocessing step that is involved in machine learning and data mining tasks for preparing data (especially...     

A novel gaussian based particle swarm optimization gravitational search algorithm for feature selection and classification

Neural Computing and Applications - A Gaussian based Particle Swarm Optimization Gravitational Search Algorithm (GPSOGSA) is being proposed for extensive feature selection that serves highly in...     

An efficient binary chaotic symbiotic organisms search algorithm approaches for feature selection problems

The Journal of Supercomputing - Feature selection is one of the main steps in preprocessing data in machine learning, and its goal is to reduce features by removing additional and noisy features....     

Time series prediction using PSO-optimized neural network and hybrid feature selection algorithm for IEEE load data

Artificial neural networks have been widely used in time series prediction. In this paper, multi-layer feedforward neural networks with optimized structures, using particle swarm optimization (PSO) algorithm, are used for hourly load prediction based on load time series of IEEE Reliability Test System. To have a small and appropriate feature subset, a hybrid method is used for feature selection in this paper. This hybrid method uses the combination of genetic algorithm (GA) and ant colony optimization (ACO) algorithm. The season, day of the week, time of the day and history load are considered as load influencing factors in this study based on the mentioned standard load dataset. The optimized number of neurons in the hidden layers of multi-layer perceptron (MLP) is determined using PSO algorithm. Experimental results show that the proposed hybrid model offers superior performance, in terms of mean absolute percentage error (MAPE), in time series prediction as compared to some recent researches in this field.     

A new feature selection algorithm and composite neural network for electricity price forecasting

In a competitive electricity market, the forecasting of energy prices is an important activity for all the market participants either for developing bidding strategies or for making investment decisions. In this paper, a new forecast strategy is proposed for day ahead prediction of electricity price, which is a complex signal with nonlinear, volatile and time dependent behavior. Our forecast strategy includes a new two stage feature selection algorithm, a composite neural network (CNN) and a few auxiliary predictors. The feature selection algorithm has two filtering stages to remove irrelevant and redundant candidate inputs, respectively. This algorithm is based on mutual information (MI) criterion and selects the input variables of the CNN among a large set of candidate inputs. The CNN is composed of a few neural networks (NN) with a new data flow among its building blocks. The CNN is the forecast engine of the proposed strategy. A kind of cross-validation technique is also presented to fine-tune the adjustable parameters of the feature selection algorithm and CNN. Moreover, the proposed price forecast strategy is equipped with a few auxiliary predictors to enrich the candidate set of inputs of the forecast engine. The whole proposed strategy is examined on the PJM, Spanish and Californian electricity markets and compared with some of the most recent price forecast methods.     

Adaptive recurrent neural network control using a structure adaptation algorithm

This paper proposes an adaptive recurrent neural network control (ARNNC) system with structure adaptation algorithm for the uncertain nonlinear systems. The developed ARNNC system is composed of a neural controller and a robust controller. The neural controller which uses a self-structuring recurrent neural network (SRNN) is the principal controller, and the robust controller is designed to achieve L ² tracking performance with desired attenuation level. The SRNN approximator is used to online estimate an ideal tracking controller with the online structuring and parameter learning algorithms. The structure learning possesses the ability of both adding and pruning hidden neurons, and the parameter learning adjusts the interconnection weights of neural network to achieve favorable approximation performance. And, by the L ² control design technique, the worst effect of approximation error on the tracking error can be attenuated to be less or equal to a specified level. Finally, the proposed ARNNC system with structure adaptation algorithm is applied to control two nonlinear dynamic systems. Simulation results prove that the proposed ARNNC system with structure adaptation algorithm can achieve favorable tracking performance even unknown the control system dynamics function.     

Correction to: An improved binary sparrow search algorithm for feature selection in data classification

Neural Computing and Applications -     

Community detection in complex networks using network embedding and gravitational search algorithm

Journal of Intelligent Information Systems - The structural and functional characteristic features of nodes can be analyzed by visualizing community structure in complex networks. Community...     

Parameter determination and feature selection for C4.5 algorithm using scatter search approach

The C4.5 decision tree (DT) can be applied in various fields and discovers knowledge for human understanding. However, different problems typically require different parameter settings. Rule of thumb or trial-and-error methods are generally utilized to determine parameter settings. However, these methods may result in poor parameter settings and unsatisfactory results. On the other hand, although a dataset can contain numerous features, not all features are beneficial for classification in C4.5 algorithm. Therefore, a novel scatter search-based approach (SS + DT) is proposed to acquire optimal parameter settings and to select the beneficial subset of features that result in better classification results. To evaluate the efficiency of the proposed SS + DT approach, datasets in the UCI (University of California, Irvine) Machine Learning Repository are utilized to assess the performance of the proposed approach. Experimental results demonstrate that the parameter settings for the C4.5 algorithm obtained by the SS + DT approach are better than those obtained by other approaches. When feature selection is considered, classification accuracy rates on most datasets are increased. Therefore, the proposed approach can be utilized to identify effectively the best parameter settings for C4.5 algorithm and useful features.     

MIMR-DGSA: Unsupervised hyperspectral band selection based on information theory and a modified discrete gravitational search algorithm

Band selection plays an important role in hyperspectral data analysis as it can improve the performance of data analysis without losing information about the constitution of the underlying data. We propose a MIMR-DGSA algorithm for band selection by following the Maximum-Information-Minimum-Redundancy (MIMR) criterion that maximises the information carried by individual features of a subset and minimises redundant information between them. Subsets are generated with a modified Discrete Gravitational Search Algorithm (DGSA) where we definine a neighbourhood concept for feature subsets. A fast algorithm for pairwise mutual information calculation that incorporates variable bandwidths of hyperspectral bands called VarBWFastMI is also developed. Classification results on three hyperspectral remote sensing datasets show that the proposed MIMR-DGSA performs similar to the original MIMR with Clonal Selection Algorithm (CSA) but is computationally more efficient and easier to handle as it has fewer parameters for tuning.     

Fuzzy-based hunger games search algorithm for global optimization and feature selection using medical data

Neural Computing and Applications - Feature selection (FS) is one of the basic data preprocessing steps in data mining and machine learning. It is used to reduce feature size and increase model...     

Transformation radial basis neural network for relevant feature selection

Radial basis neural networks are excellent candidates for selecting relevant features in pattern recognition problems. By a slight change in the traditional three-layer architecture of a radial basis neural network, we can obtain a quantitative method, which allows us to get a ranking within the features. We present a new neural network concept, combining at the same time two different skills: classification and detection of relevant features in the input vector.     

Efficient text independent speaker recognition with wavelet feature selection based multilayered neural network using supervised learning algorithm

A wavelet packet feature selection derived by using multilayered neural network for speaker identification is described. The concept of a multilayered neural network is without using a gradient method. First, the outputs of each hidden unit are algebraically determined by an error backpropagation method. Then, the weight parameters are determined by using an exponentially weighted least squares method. Our results have shown that this feature selection introduced better performance than the other methods with respect to the percentages of recognition.     

An information criterion for optimal neural network selection

The choice of an optimal neural network design for a given problem is addressed. A relationship between optimal network design and statistical model identification is described. A derivative of Akaike's information criterion (AIC) is given. This modification yields an information statistic which can be used to objectively select a ;best' network for binary classification problems. The technique can be extended to problems with an arbitrary number of classes.