Genetic Algorithm For Feature Selection Research Articles

Extracting multi-features and optimizing feature space with sparse auto-encoder over WorldView-2 images

ABSTRACTFeature selection of very high-resolution (VHR) images is a key prerequisite for supervised classification. However, it is always difficult to acquire the features which have the highest correlation to the type of land cover for improving classification accuracy. To address this problem, this paper proposed a methodology of feature selection using the results of multiple segmentation via genetic algorithm (GA) and correlation feature selection (CFS) integrating sparse auto-encoder (SAE). Firstly, 61 features, including spectral features and spatial features, are extracted from the results of multi-scale segmentation over a WorldView-2 image in Xicheng District, Beijing. Then, 40-dimensional features and 30-dimensional features are derived from the selection with GA+CFS and the optimization with SAE, respectively. Thirdly, the final classification is achieved by logistic regression (LR) based on different subsets of features extracted from the WorldView-2 image. It is found that the result of feature selection could contribute to increase in the intra-species separation and reduction in the inner-species variability. Adding extra lower-ranked features appeared to reduce the accuracy of classification. The results indicate that the overall classification accuracy with 30-dimensional features reached 87.56%, and increased 5.61% compared to the results with 61-dimensional features. For the two kinds of optimized features, the Z-test values are all greater than 1.96, which implied that feature dimensionality reduction and feature space optimization could significantly improve the accuracy of image land cover classification. The texture features in the wavelet domain are the most important features for the study area in the WorldView-2 image classification. Adding wavelet and the grey-level co-occurrence matrix (GLCM) information, especially for GLCM features in wavelet, appeared not to improve classification accuracy. The SAE-based method can produce feature subsets for improving mapping accuracy more efficiently.

Deluge based Genetic Algorithm for feature selection

Feature selection methods are used to identify and remove irrelevant and redundant attributes from the original feature vector that do not have much contribution to enhance the performance of a predictive model. Meta-heuristic feature selection algorithms, used as a solution to this problem, need to have a good trade-off between exploitation and exploration of the search space. Genetic Algorithm (GA), a popular meta-heuristic algorithm, lacks exploitation capability, which in turn affects the local search ability of the algorithm. Basically, GA uses mutation operation to take care of exploitation which has certain limitations. As a result, GA gets stuck in local optima. To encounter this problem, in the present work, we have intelligently blended the Great Deluge Algorithm (GDA), a local search algorithm, with GA. Here GDA is used in place of mutation operation of the GA. Application of GDA yields a high degree of exploitation through the use of perturbation of candidate solutions. The proposed method is named as Deluge based Genetic Algorithm (DGA). We have applied the DGA on 15 publicly available standard datasets taken from the UCI dataset repository. To show the classifier independent nature of the proposed feature selection method, we have used 3 different classifiers namely K-Nearest Neighbour (KNN), Multi-layer Perceptron (MLP) and Support Vector Machine (SVM). Comparison of DGA has been performed with other contemporary algorithms like the basic version of GA, Particle Swarm Optimisation (PSO), Simulated Annealing (SA) and Histogram based Multi-Objective GA (HMOGA). From the comparison results, it has been observed that DGA performs much better than others in most of the cases. Thus, our main contributions in this paper are introduction of a new variant of GA for FS which uses GDA to strengthen its exploitational ability and application of the proposed method on 15 well-known UCI datasets using KNN, MLP and SVM classifiers.

A class-based approach to classify PolSAR imagery using optimum classifier

ABSTRACT Classification of Polarimetric SAR (PolSAR) imagery is still one of the challenges in active remote sensing applications. Although a large number of features and different classifiers have been proposed, no unique approach has been found to satisfy all of the images and classes yet. In this paper, given the extracted features from PolSAR data, a new class-based feature selection (CBFS) algorithm is proposed to find the most optimum features for each class. Maximizing the discrimination of each class from the others is the main contribution of the CBFS which yields distinctive features. The selected features are then employed by classifiers to generate different classification results. Finally, a new approach is developed to combine these classification results to produce the final land cover map. Five different classifiers of Wishart Maximum Likelihood, Gaussian Maximum Likelihood, Support Vector Machine, Multi Layer Perceptron and Fuzzy Inference System are also used for classification. Given the CBFS results, two different Radarsat-2 and AirSAR PolSAR data were classified. Selected features led to improvement of about 5% in producer accuracies in comparison with two well-known Genetic Algorithm Feature Selection (GAFS) and Prototype Space Feature Selection (PSFS) methods. Moreover, Comparison results demonstrate that the fuzzy classifiers could improve the accuracies about 3% if they are suitably constructed and well designed. The achieved higher overall accuracy for the final classified map shows the effectiveness of the proposed approach over the other compared classification procedures.

A Nested Genetic Algorithm for feature selection in high-dimensional cancer Microarray datasets

Cancer is a dangerous disease that causes death worldwide. Discovering few genes relevant to one cancer disease can result in effective treatments. The challenge associated with the Microarray datasets is its high dimensionality; the huge number of features compared to the modest number of samples in these datasets. Recent research efforts attempted to reduce this high-dimensionality using different feature selection techniques. This paper presents an ensemble feature selection technique based on t-test and genetic algorithm. After preprocessing the data using t-test, a Nested Genetic Algorithm, namely Nested-GA, is used to get the optimal subset of features by combining data from two different datasets. Nested-GA consists of two Nested Genetic Algorithms (outer and inner) that run on two different kinds of datasets. The Outer Genetic Algorithm (OGA-SVM) works on Microarray gene expression datasets, whereas the Inner Genetic Algorithm (IGA-NNW) runs on DNA Methylation datasets. Nested-GA is performed on a colon cancer dataset with 5-fold cross validation. After applying Nested-GA, the Incremental Feature Selection (IFS) strategy is used to get the smallest optimal genes subset. The genes subset has been validated on an independent dataset resulting in 99.9% classification accuracy. Consequently, the biological significance of the resulting optimal genes is validated using Enrichment Analysis. Moreover, the results of Nested-GA have been compared to the results of other feature selection algorithms that have been run on either Gene Expression or DNA Methylation datasets. From the experimental results, Nested-GA showed the highest classification performance with a small optimal feature subset compared to the other algorithms. Furthermore, by running Nested-GA on lung cancer datasets that contain two different cancer subtypes, it resulted in significantly better classification accuracy (98.4%) compared to the accuracy of a previous research (84.6%) that utilized lung cancer DNA-Methylation data only.

A Genetic-LVQ neural networks approach for handwritten Arabic character recognition

Handwritten recognition systems are a dynamic field of research in areas of artificial intelligence. Many smart devices available in the market such as pen-based computers, tablets, mobiles with handwritten recognition technology need to rely on efficient handwritten recognition systems. In this paper we present a novel Arabic character handwritten recognition system based on a hybrid method consisting of a genetic algorithm and a Learning vector quantization (LVQ) neural network. Sixty different handwritten Arabic character datasets are used for training the neural network. Each character dataset contains 28 letters written twice with 15 distinct shaped alphabets, and each handwritten Arabic letter is represented by a binary matrix that is used as an input to a genetic algorithm for feature selection and dimension reduction to include only the most effective features to be fed to the LVQ classifier. The recognition process in the system involves several essential steps such as: handwritten letter acquisition, dataset preparation, feature selection, training, and recognition. Comparing our results to those acquired by the whole feature dataset without selection, and to the results using other classification algorithms confirms the effectiveness of our proposed handwritten recognition system with an accuracy of 95.4%, hence, showing a promising potential for improving future handwritten Arabic recognition devices in the market.

Diagnosis of mesothelioma with deep learning.

Malignant mesothelioma (MM) is a rare but aggressive cancer. The definitive diagnosis of MM is critical for effective treatment and has important medicolegal significance. However, the definitive diagnosis of MM is challenging due to its composite epithelial/mesenchymal pattern. The aim of the current study was to develop a deep learning method to automatically diagnose MM. A retrospective analysis of 324 participants with or without MM was performed. Significant features were selected using a genetic algorithm (GA) or a ReliefF algorithm performed in MATLAB software. Subsequently, the current study constructed and trained several models based on a backpropagation (BP) algorithm, extreme learning machine algorithm and stacked sparse autoencoder (SSAE) to diagnose MM. A confusion matrix, F-measure and a receiver operating characteristic (ROC) curve were used to evaluate the performance of each model. A total of 34 potential variables were analyzed, while the GA and ReliefF algorithms selected 19 and 5 effective features, respectively. The selected features were used as the inputs of the three models. SSAE and GA+SSAE demonstrated the highest performance in terms of classification accuracy, specificity, F-measure and the area under the ROC curve. Overall, the GA+SSAE model was the preferred model since it required a shorter CPU time and fewer variables. Therefore, the SSAE with GA feature selection was selected as the most accurate model for the diagnosis of MM. The deep learning methods developed based on the GA+SSAE model may assist physicians with the diagnosis of MM.

Linear Discriminant Analysis Based Genetic Algorithm with Generalized Regression Neural Network – A Hybrid Expert System for Diagnosis of Diabetes

Among the applications enabled by expert systems, disease diagnosis is a particularly important one. Nowadays, diabetes is found to be a complex health issue in human life. There has been a wide range of intelligent methods proposed for early detection of diabetes. The objective of this paper is to propose an expert system for better diagnosis of diabetes. The methodology of the proposed framework is classified as two stages: (a) Linear Discriminant Analysis (LDA) based genetic algorithm for feature selection, (b) Generalized Regression Neural Network (GRNN) for classification. The proposed a genetic algorithm with Linear Discriminant Analysis (LDA) based feature selection for not only reduce the computation time and cost of the disease diagnosis but also improved the accuracy of classification. The performance of the method is evaluated using the calculation of accuracy, confusion matrix and Receiver-Operating Characteristic (ROC). The proposed method is compared with other existing methods for evaluating the performance and accuracy. The LDA based Genetic Algorithm (GA) with GRNN produces the accuracy of 80.2017% with a ROC of 0.875.

Soil salinity mapping using dual-polarized SAR Sentinel-1 imagery

ABSTRACTSoil salinity is a major environmental threat, which has a negative impact on soil productivity and agricultural fields. One of the most promising methods for monitoring affected areas, which has special importance in land management studies, is through remote sensing technologies. While the potential of optical imagery in detecting saline soils is widely investigated, limited studies have been dedicated to assessing the potential of Synthetic Aperture Radar (SAR) imagery in monitoring soil salinity. Accordingly, this paper deals with soil salinity estimation using Sentinel-1 SAR imagery in an area which is highly affected by salinity hazard. Due to lack of a suitable theoretical model for simulating radar backscatter of soil based on salt contents, we investigated a new method to relate radar intensity to measured in-situ salinity directly. In the first step, Sentinel-1 dual polarized VV (Vertical transmit Vertical receive) and VH (Vertical transmit Horizontal receive) data were acquired from the study site. A field study was also conducted, simultaneously, and the Electrical Conductivity (EC) of several soil samples was measured. We then extracted some features based on the intensity images of both VV and VH polarization. Based on the fact that the target texture affects the radar response, an analysis of the texture was also performed by calculating the first and second order statistics, extracted from the histogram and Gray Level Co-occurrence Matrix (GLCM), respectively. The Support Vector Regression (SVR) technique, with different kernel functions, was used to relate explanatory variables to ground measured salinity. We also applied Feature Selection (FS) algorithms of the Genetic Algorithm (GA) and Sequential Feature Selection (SFS) for optimizing the model and selecting the best explanatory features. The results showed that ε-SVR with Radial Basis Function (RBF) kernel had the most accuracy with the Coefficient of Determination (R2) = 0.9783 and Root Mean Square Error (RMSE) = 0.3561 when the GA FS was applied. Also and had the best performance in salinity detection. It can be concluded that the intensity images of VV and VH polarization of SAR imagery have the potential to discriminate saline surface soils, regardless of the failure of common backscattering models.

Multi-stage SVM approach for cardiac arrhythmias detection in short single-lead ECG recorded by a wearable device

Objective: Use of wearable ECG devices for arrhythmia screening is limited due to poor signal quality, small number of leads and short records, leading to incorrect recognition of pathological events. This paper introduces a novel approach to classification (normal/‘N’, atrial fibrillation/‘A’, other/‘O’, and noisy/‘P’) of short single-lead ECGs recorded by wearable devices. Approach: Various rhythm and morphology features are derived from the separate beats (‘local’ features) as well as the entire ECGs (‘global’ features) to represent short-term events and general trends respectively. Various types of atrial and ventricular activity, heart beats and, finally, ECG records are then recognised by a multi-level approach combining a support vector machine (SVM), decision tree and threshold-based rules. Main results: The proposed features are suitable for the recognition of ‘A’. The method is robust due to the noise estimation involved. A combination of radial and linear SVMs ensures both high predictive performance and effective generalisation. Cost-sensitive learning, genetic algorithm feature selection and thresholding improve overall performance. The generalisation ability and reliability of this approach are high, as verified by cross-validation on a training set and by blind testing, with only a slight decrease of overall F1-measure, from 0.84 on training to 0.81 on the tested dataset. ‘O’ recognition seems to be the most difficult (test F1-measures: 0.90/‘N’, 0.81/‘A’ and 0.72/‘O’) due to high inter-patient variability and similarity with ‘N’. Significance: These study results contribute to multidisciplinary areas, focusing on creation of robust and reliable cardiac monitoring systems in order to improve diagnosis, reduce unnecessary time-consuming expert ECG scoring and, consequently, ensure timely and effective treatment.

Development of Spectral Disease Indices for 'Flavescence Dorée' Grapevine Disease Identification.

Spectral measurements are employed in many precision agriculture applications, due to their ability to monitor the vegetation’s health state. Spectral vegetation indices are one of the main techniques currently used in remote sensing activities, since they are related to biophysical and biochemical crop variables. Moreover, they have been evaluated in some studies as potentially beneficial for detecting or differentiating crop diseases. Flavescence Dorée (FD) is an infectious, incurable disease of the grapevine that can produce severe yield losses and, hence, compromise the stability of the vineyards. The aim of this study was to develop specific spectral disease indices (SDIs) for the detection of FD disease in grapevines. Spectral signatures of healthy and diseased grapevine leaves were measured with a non-imaging spectro-radiometer at two infection severity levels. The most discriminating wavelengths were selected by a genetic algorithm (GA) feature selection tool, the Spectral Disease Indices (SDIs) are designed by exhaustively testing all possible combinations of wavelengths chosen. The best weighted combination of a single wavelength and a normalized difference is chosen to create the index. The SDIs are tested for their ability to differentiate healthy from diseased vine leaves and they are compared to some common set of Spectral Vegetation Indices (SVIs). It was demonstrated that using vegetation indices was, in general, better than using complete spectral data and that SDIs specifically designed for FD performed better than traditional SVIs in most of cases. The precision of the classification is higher than 90%. This study demonstrates that SDIs have the potential to improve disease detection, identification and monitoring in precision agriculture applications.

Distance Based Genetic Algorithm for Feature Selection in Computer Aided Diagnosis Systems

Introduction: One of the most challenging tasks in computer-aided diagnosis (CAD) is mapping of radiological features to image descriptors. If the descriptors chosen are not appropriate to discriminate a disorder from another, the role of CAD system may not be appreciable. Keywords: Computer aided diagnosis, feature selection, genetic algorithm, k-nearest neighbor classifier, lung disorder, objective function.

A hybrid system with filter approach and multiple population genetic algorithm for feature selection in credit scoring

With the financial crisis happened in 2007, massive credit risks are exposed to the banking sectors. So credit scoring has attracted more and more attention. Bank owns a lot of customer data. By using those data, credit scoring model can judge the applicants’ credit risk accurately. But those data are often high dimensional, and have some irrelevant features. Those irrelevant features will affect classifiers accuracy. Therefore, feature selection is an important topic. This paper proposes a two-phase hybrid approach based on filter approach and multiple population genetic algorithm-HMPGA. In phase 1, it introduces the idea of wrapper approach into three filter approaches to acquire some important prior information for initial populations setting of MPGA. In phase 2, it takes advantage of MPGA’s characteristics of global optimization and quick convergence to find optimal feature subset. This paper uses two real credit scoring datasets of UCI databases to compare HMPGA, MPGA and GA. It verifies that the accuracies of feature subsets acquired from HMPGA, MPGA and GA are superior to three filter approaches. Meanwhile, nonparametric Wilcoxon signed rank test is held to confirm that HMPGA is better than MPGA and GA. HMPGA not only can be applied to feature selection of credit scoring, but also can be applied to more fields of data mining.

A tribe competition-based genetic algorithm for feature selection in pattern classification

Feature selection has always been a critical step in pattern recognition, in which evolutionary algorithms, such as the genetic algorithm (GA), are most commonly used. However, the individual encoding scheme used in various GAs would either pose a bias on the solution or require a pre-specified number of features, and hence may lead to less accurate results. In this paper, a tribe competition-based genetic algorithm (TCbGA) is proposed for feature selection in pattern classification. The population of individuals is divided into multiple tribes, and the initialization and evolutionary operations are modified to ensure that the number of selected features in each tribe follows a Gaussian distribution. Thus each tribe focuses on exploring a specific part of the solution space. Meanwhile, tribe competition is introduced to the evolution process, which allows the winning tribes, which produce better individuals, to enlarge their sizes, i.e. having more individuals to search their parts of the solution space. This algorithm, therefore, avoids the bias on solutions and requirement of a pre-specified number of features. We have evaluated our algorithm against several state-of-the-art feature selection approaches on 20 benchmark datasets. Our results suggest that the proposed TCbGA algorithm can identify the optimal feature subset more effectively and produce more accurate pattern classification.

Distance Based Genetic Algorithm for Feature Selection in Computer Aided Diagnosis Systems

Distance Based Genetic Algorithm for Feature Selection in Computer Aided Diagnosis Systems

Cloud computing-based parallel genetic algorithm for gene selection in cancer classification

Cancer classification is one of the main steps during patient healing process. This fact enforces modern clinical researchers to use advanced bioinformatics methods for cancer classification. Cancer classification is usually performed using gene expression data gained in microarray experiment and advanced machine learning methods. Microarray experiment generates huge amount of data, and its processing via machine learning methods represents a big challenge. In this study, two-step classification paradigm which merges genetic algorithm feature selection and machine learning classifiers is utilized. Genetic algorithm is built in MapReduce programming spirit which makes this algorithm highly scalable for Hadoop cluster. In order to improve the performance of the proposed algorithm, it is extended into a parallel algorithm which process on microarray data in distributed manner using the Hadoop MapReduce framework. In this paper, the algorithm was tested on eleven GEMS data sets (9 tumors, 11 tumors, 14 tumors, brain tumor 1, lung cancer, brain tumor 2, leukemia 1, DLBCL, leukemia 2, SRBCT, and prostate tumor) and its accuracy reached 100% for less than 25 selected features. The proposed cloud computing-based MapReduce parallel genetic algorithm performed well on gene expression data. In addition, the scalability of the suggested algorithm is unlimited because of underlying Hadoop MapReduce platform. The presented results indicate that the proposed method can be effectively implemented for real-world microarray data in the cloud environment. In addition, the Hadoop MapReduce framework demonstrates substantial decrease in the computation time.

A Hybrid Binary Cuckoo Search and Genetic Algorithm for Feature Selection in Type-2 Diabetes

Data mining techniques are applied in bioinformatics to analyze biomedical data. When the number of features related to the data is irrelevant, the classifiers will produce unsatisfactory results. This paper addresses the need to analyze the data for extracting relevant features. A number of feature selection algorithms are developed in the field of medical data to address feature selection. In this paper, an intelligent hybrid optimal feature selection algorithm is proposed for Type-2 diabetes with improved classification accuracy. This work proposes an intelligent Hybrid Binary Cuckoo Search (CS) and Genetic Algorithm (GA) for selecting the important features of Type-2 diabetes. In HBCS-GA, exploration and exploitation of CS is improved using genetic operators to select relevant features with better accuracy. To validate the model, a 10-fold cross-validation strategy is used. The proposed algorithm produces 99.31% accuracy to diagnose the disease. The performance of HBCS-GAis also compared with other approaches. Also, the model validation with reduced features is performed with Decision Tree(DT), Bayesian Network(BN), Support Vector Machine (SVM) and k-Nearest Neighbor (k-NN) classifiers. The accuracies obtained are 94.46%, 96.07%, 98.84%, 96.79% respectively. The results also showed that HBCS-GA achieved high classification accuracy than the other approaches. Keywords: Diabetes, normalization, classification, instance selection, wrapper feature selection, cuckoo search, genetic algorithm.

Performance Analysis of GA and PSO based Feature Selection Techniques for Improving Classification Accuracy in Remote Sensing Images

Background/Objectives: - Feature Selection is applicable to decrease the number of features in various applications wherein the data include hundreds and thousands of features. The objective of this study is to choose Genetic Algorithm for feature selection to obtain better fitness function. Methods/Statistical Analysis: Particle Swarm Optimization (PSO) approach is used for selecting the subset from the combination of texture based features and providing the better fitness values. In this paper PSO is used to obtain the feature sets and the performance is compared with genetic algorithm. Support vector machine classifier is used to improve the classification accuracy. Findings: The experimental results shows that PSO overall accuracy is improved to LISS IV 1.7%, 1.4% and 2.9% and the kappa coefficient is improved to 0.06%, 0.012% and 0.39% as compared to GA. Application/Improvements: The Fitness value obtained by GA is more complex and not accurate. To reduce the complexity and increase the accuracy Particle Swarm optimization is used. Hence PSO improved the quality of texture based images.

Dynamic box office forecasting based on microblog data

Movies, as one of the most rapidly developing industries? outcomes, have gained much attention these years. Especially in China, the world? s second largest film market with a rapid growing speed, many film companies intend to foresee the future box office in advance to better arrange their income and expenditure. Unlike some traditional forecasting model based on several movie-related features, this paper comprehensively utilizes the real-time social media, microblog, to realize a more accurate weekly box office forecasting model. The features weekly extracted from microblogs can be divided into count based features and context based features, along with the existing box office and the screen arrangements, to predict the box office in next week. For count based features, not only the total volume of related microblogs and the diffusion effect considers the number of followers, several unnoticed features like authentication users, gender ratio and mobile-users ratio are also introduced into the original predicting model. For content based features, a duplicate semantic analysis method is proposed. The number of tweets which can indeed influence others? purchase decision, along with the number of tweets with positive and negative influence is the results of the analysis system. On this basis, guided effect for each influential tweets are identified by the praise, comment and retweet times. Some machine learning models are then adopted after using genetic algorithm (GA) for feature selection. The empirical study shows that our research can dynamic forecast box office with a sustainable good performance.

An Image based Diagnostic System for Lung Disease Classification

Model-based detection and classification of nodules are two major steps in CAD systems design and evaluation. A common health problem, lung diseases are the most prevailing medical conditions throughout the world. In this paper, Lung diseases are automatically classified as Emphysema, Bronchitis, Pleural effusion and normal lung.The lung CT images are taken as input, preprocessing is applied, feature extraction is done by various methods such as Gabor filter extracts the texture features, walsh hadamard transform extracts the pixel co-efficient values, and a fusion method is proposed in this work which extracts the median absolute deviation values. Feature selection including statistical correlation based methods and Genetic Algorithm for searching in feature vector space are investigated. Four types of the classifiers are used where the Multi-Layer Perceptron Neural Network (MLP-NN) classifier with proposed fusion feature extraction method, genetic algorithm feature selection method gives promising result of 91% accuracy than J48, K- Nearest Neighbour and Naïve bayes classifiers.

A hybrid method for classifying cognitive states from fMRI data.

Functional magnetic resonance imaging (fMRI) makes it possible to detect brain activities in order to elucidate cognitive-states. The complex nature of fMRI data requires under-standing of the analyses applied to produce possible avenues for developing models of cognitive state classification and improving brain activity prediction. While many models of classification task of fMRI data analysis have been developed, in this paper, we present a novel hybrid technique through combining the best attributes of genetic algorithms (GAs) and ensemble decision tree technique that consistently outperforms all other methods which are being used for cognitive-state classification. Specifically, this paper illustrates the combined effort of decision-trees ensemble and GAs for feature selection through an extensive simulation study and discusses the classification performance with respect to fMRI data. We have shown that our proposed method exhibits significant reduction of the number of features with clear edge classification accuracy over ensemble of decision-trees.