Feature Reduction Process Research Articles

Optimized Radial Basis Neural Network for Classification of Breast Cancer Images.

Breast cancer is a curable disease if diagnosed at an early stage. The chances of having breast cancer are the lowest in married women after the breast-feeding phase because the cancer is formed from the blocked milk ducts. Nowadays, cancer is considered the leading cause of death globally. Breast cancer is the most common cancer among females. It is possible to develop breast cancer while breast-feeding a baby, but it is rare. Mammography is one of the most effective methods used in hospitals and clinics for early detection of breast cancer. Various researchers are used in artificial intelligence- based mammogram techniques. This process of mammography will reduce the death rate of the patients affected by breast cancer. This process is improved by the image analysing, detection, screening, diagnosing, and other performance measures. The radial basis neural network will be used for classification purposes. The radial basis neural network is designed with the help of the optimization algorithm. The optimization is to tune the classifier to reduce the error rate with the minimum time for the training process. The cuckoo search algorithm will be used for this purpose. Thus, the proposed optimum RBNN is determined to classify breast cancer images. In this, the three sets of properties were classified by performing the feature extraction and feature reduction. In this breast cancer MRI image, the normal, benign, and malignant is taken to perform the classification. The minimum fitness value is determined to evaluate the optimum value of possible locations. The radial basis function is evaluated with the cuckoo search algorithm to optimize the feature reduction process. The proposed methodology is compared with the traditional radial basis neural network using the evaluation parameter like accuracy, precision, recall and f1-score. The whole system model is done by using Matrix Laboratory (MATLAB) with the adaptation of 2018a since the proposed system is most efficient than most recent related literature. Thus, it concluded with the efficient classification process of RBNN using a cuckoo search algorithm for breast cancer images. The mammogram images are taken into recent research because breast cancer is a major issue for women. This process is carried to classify the various features for three sets of properties. The optimized classifier improves performance and provides a better result. In this proposed research work, the input image is filtered using a wiener filter, and the classifier extracts the feature based on the breast image.

Stock Market Prediction Framework Based on the Fruit Fly Optimization Method

Stock market prediction helps investors in decision-making process of investment to achieve profit. Recently, the deep learning method shows the significant performance in the stock market prediction. These deep learning models have the drawback of overfitting problems when it processing number of features. In this research, the fruit fly optimization method has been proposed for the feature reduction process in the stock market prediction. The fruit fly method has the advantages of simple computation processes and less number of parameter for tuning. The fruit fly method selects more relevant features to reduce the overfitting problem in the Long Short Term Memory (LSTM) classifiers. The Nifty 50 and S&P 500 data were applied to test the efficiency of the proposed model. The obtained result shows that the fruit fly method based framework achieved more efficiency than other techniques. The fruit fly based framework has 0.426 of Mean Square Error (MSE) and the existing firefly method has 0.621 MSE.

MRI Radiomics for the Prediction of Fuhrman Grade in Clear Cell Renal Cell Carcinoma: a Machine Learning Exploratory Study.

The Fuhrman nuclear grade is a recognized prognostic factor for patients with clear cell renal cell carcinoma (CCRCC) and its pre-treatment evaluation significantly affects decision-making in terms of management. In this study, we aimed to assess the feasibility of a combined approach of radiomics and machine learning based on MR images for a non-invasive prediction of Fuhrman grade, specifically differentiation of high- from low-grade tumor and grade assessment. Images acquired on a 3-Tesla scanner (T2-weighted and post-contrast) from 32 patients (20 with low-grade and 12 with high-grade tumor) were annotated to generate volumes of interest enclosing CCRCC lesions. After image resampling, normalization, and filtering, 2438 features were extracted. A two-step feature reduction process was used to between 1 and 7 features depending on the algorithm employed. A J48 decision tree alone and in combination with ensemble learning methods were used. In the differentiation between high- and low-grade tumors, all the ensemble methods achieved an accuracy greater than 90%. On the other end, the best results in terms of accuracy (84.4%) in the assessment of tumor grade were achieved by the random forest. These evidences support the hypothesis that a combined radiomic and machine learning approach based on MR images could represent a feasible tool for the prediction of Fuhrman grade in patients affected by CCRCC.

Deep Learning on Digital Image Splicing Detection Using CFA Artifacts

Digital image forgery is a serious problem of an increasing attention from the research society. Image splicing is a well-known type of digital image forgery in which the forged image is synthesized from two or more images. Splicing forgery detection is more challenging when compared with other forgery types because the forged image does not contain any duplicated regions. In addition, unavailability of source images introduces no evidence about the forgery process. In this study, an automated image splicing forgery detection scheme is presented. It depends on extracting the feature of images based on the analysis of color filter array (CFA). A feature reduction process is performed using principal component analysis (PCA) to reduce the dimensionality of the resulting feature vectors. A deep belief network-based classifier is built and trained to classify the tested images as authentic or spliced images. The proposed scheme is evaluated through a set of experiments on Columbia Image Splicing Detection Evaluation Dataset (CISDED) under different scenarios including adding postprocessing on the spliced images such JPEG compression and Gaussian Noise. The obtained results reveal that the proposed scheme exhibits a promising performance with 95.05% precision, 94.05% recall, 94.05% true positive rate, and 98.197% accuracy. Moreover, the obtained results show the superiority of the proposed scheme compared to other recent splicing detection method.

Amyotrophic Lateral Sclerosis EEG Classification Using Deep Neural Network And TLBO

The prediction and detection of amyotrophic lateral sclerosis (ALS) diseases are a challenging task for the brain-computer interface. The discovery and classification of amyotrophic lateral sclerosis diseases used various neural network models. The limitation of the neural network deals with feature extraction and vector conversion of ALS data. The extraction of features of ALS diseases is very complex due to the recorded signal of the human brain. The recoded signal consists of the actual signal and other brain behaviours of a signal treat as noise. The noise recorded signal degraded the performance of the brain-computer interface. For the betterment of the brain-computer interface, I am using feature selection and reduction process with the neural network model. This paper proposed the deep learning-based classification algorithms for the prediction of ALS disease. The deep learning algorithm is a variant of a multilayer neural network for the extraction of features used discrete wavelet transform function. The discrete wavelet transform function decomposed the electroencephalogram signal in different sub-bands for the processing. The represent bands into varied frequency range. The proposed algorithms are simulated in MATLAB software and used the standard dataset of ALS diseases for the analysis of performance. The proposed algorithm compares with Bayesian-based neural network and ensemble-based machine learning classifiers for ALS disease detection. The performance of the proposed algorithms improved the efficiency of the brain-computer interface system.

Feature selection and dimensionality reduction: An extensive comparison in hand gesture classification by sEMG in eight channels armband approach

Gesture recognition by surface electromyography (sEMG) signals is used for several applications as prosthesis control and human-machines interfaces. One of trending approaches to sEMG acquisition is the multiple-channel armband with equidistant electrodes. Several efforts are made to improve the performance of these devices in gestures recognition applications, especially in feature selection. It is necessary choose one approach for feature selection due to the great number of electromyography features available to use. In this work, an extensive comparison of feature reduction techniques and their influences in the classification process is presented. Unlike other works, we presents the comparison between methods of feature selection and classification; this main contribution is show how the feature reduction process can aid and increase the performance of classification of sEMG in armband acquisition approach. Two general methods were employed, feature selection by wrapper forward stepwise and dimensionality reduction, resulting in eight different techniques. The following dimensionality reduction techniques were used: Principal Component Analysis, Linear Discriminant Analysis, Isomap, Manifold Charting, Autoencoder, t-distributed Stochastic Neighbor Embedding, and Large Margin Nearest Neighbor (LMNN). Seven classifiers were used, aiming at recognize six gestures acquired from an 8-channel armband of 13 subjects. An average accuracy of 89.4 % was obtained with 5 features and an Extreme Learning Machine classifier, in the feature selection approach. On other hand, considering 40 dimensions, an average accuracy of 94 % was obtained, regarding a combination between Support Vector Machine with Gaussian kernel and a LMNN technique. These results showed significant differences in statistical tests.

High-Performance Feature Selection Model for Network Intrusion Detection System

Network intrusions detection is a continuous vigilant task and to efficiently analyze the traffic in the corporate network to detect network intrusions. The efficiency of the Network Intrusion Detection System (NIDS) performance can be improved by adopting feature selection or reduction process to suit the present day high speed real time networks. This work is focused on identifying the key features of the audit dataset used to build an efficient light-weight NIDS. The NSL KDD dataset is used in this work titled Attribute Richness Based Feature Selection (ARFS) in order to analyze its performance.The obtained results are compared with the Correlation-based Feature Selection (CFS) and Information Gain (IG) feature selection methods. The proposed feature selection method produced better detection rate comparatively.

A machine learning system for carotid plaque vulnerability assessment based on ultrasound images

Purpose/Background/ObjectivesCarotid plaque vulnerability assessment is essential for the identification of high-risk patients. A specific mouse model for the study of carotid atherosclerosis has been recently developed. Aim of this study was to develop a predictive mathematical model for carotid plaque vulnerability assessment based on the post processing of micro-Ultrasound (mUS) images only.Methods17 ApoE-/- male mice (16 weeks) were employed. After three weeks of high-fat diet, a tapered cast, designed to induce the formation of an unstable plaque upstream from the cast and a stable one downstream from it, was surgically placed around the right common carotid. mUS examination was repeated before the surgical procedure and after three months from it. Color-Doppler, B-mode and Pulsed-wave Doppler images were acquired to assess morphological, functional and hemodynamic parameters. In particular, texture analysis was applied on both the atherosclerotic lesions post-processing B-mode images. Peak velocity (Vp), Relative Turbolence Intensity (rTI) and velocity range (rangevel) were assessed from PW-Doppler images. Relative Distension (relD) and Pulse Wave Velocity (PWV) were evaluated for both the regions. All the mUS indexes underwent a feature reduction process and were used to train different machine learning approaches.ResultsThe downstream region presented higher PWV values than the upstream one; furthermore, it was characterized by higher values of rTI and rangevel. The weighted kNN classifier supplied the best providing 92.6% accuracy, 91% sensitivity and 94% specificity.ConclusionsThe mathematical predictive model could represent a valid approach to be translated in the clinical field and easily employed in clinical practice.

Feature Reduction Based on Hybrid Efficient Weighted Gene Genetic Algorithms with Artificial Neural Network for Machine Learning Problems in the Big Data

A large amount of data being generated from different sources and the analyzing and extracting of useful information from these data becomes a very complex task. The difficulty of dealing with big data arises from many factors such as the high number of features, existence of lost data, and variety of data. One of the most effective solutions that used to overcome the huge amount of big data is the feature reduction process. In this paper, a set of hybrid and efficient algorithms are proposed to classify the datasets that have large feature size by merging the genetic algorithms with the artificial neural networks. The genetic algorithms are used as a prestep to significantly reduce the feature size of the analyzed data before handling that data using machine learning techniques. Reducing the number of features simplifies the task of classifying the analyzed data and enhances the performance of the machine learning algorithms that are used to extract valuable information from big data. The proposed algorithms use a new gene-weight mechanism that can significantly enhance the performance and decrease the required search time. The proposed algorithms are applied on different datasets to pick the most relative and important features before applying the artificial neural networks algorithm, and the results show that our proposed algorithms can effectively enhance the classifying performance over the tested datasets.

Sensor selection scheme in activity recognition based on hierarchical feature reduction

To better understand the activity state of human, we might need multiple sensors on different parts of the body. According to different types of activities, the number and slot of required sensors would also be different. Therefore, how to determine the number and slot of necessary sensors regarding to wearers’ experience and processing efficiency is a meaningful study in actual practice. In this work, we propose a novel sensor selection scheme that is based on the improvement of the feature reduction process of the recognition. This scheme applies a hierarchical feature reduction method based on mutual information with max relevance and low-dimensional embedding strategy. It divides the process of feature reduction into two stages: first, redundant sensors are removed with one-order sequential forward selection based on mutual information; second, feature selection strategy that maximizing class-relevance is integrated with low-dimensional mapping so that the set of features will be further compressed. To verify the feasibility and superiority of the scheme, we design a complete solution for real practice of human activity recognition. According to the results of the experiments, we are able to recognize human activities accurately and efficiently with as few sensors as possible.

Hierarchical Feature Reduction with Max Relevance and Low Dimensional Embedding Strategy and Its Application in Activity Recognition with Multi-sensors

Human activity recognition is widely discussed in many domains, and wearable sensors have proved to be a wise choice in related studies. Regarding activity estimation based on multiple sensors, we focus on the feature reduction process based on wearers’ experience and processing efficiency. According to the problem of determining the number and positions of necessary sensors in actual practice, we propose a hierarchical feature reduction method based on mutual information with max relevance and low-dimensional embedding strategies. This method divides the process of feature reduction into two stages: firstly, redundant sensors are eliminated with one-order sequential forward selection based on mutual information; secondly, feature selection strategy that maximizing class-relevance is integrated with low dimensional mapping so that the set of features will be further compressed.

Domain-aware Mashup service clustering based on LDA topic model from multiple data sources

ContextMashup is emerging as a promising software development method for allowing software developers to compose existing Web APIs to create new or value-added composite Web services. However, the rapid growth in the number of available Mashup services makes it difficult for software developers to select a suitable Mashup service to satisfy their requirements. Even though clustering based Mashup discovery technique shows a promise of improving the quality of Mashup service discovery, Mashup service clustering with high accuracy and good efficiency is still a challenge problem. ObjectiveThis paper proposes a novel domain-aware Mashup service clustering method with high accuracy and good efficiency by exploiting LDA topic model built from multiple data sources, to improve the quality of Mashup service discovery. MethodThe proposed method firstly designs a domain-aware Mashup service feature selection and reduction process by refining characterization of their domains to consolidate domain relevance. Then, it presents an extended LDA topic model built from multiple data sources (include Mashup description text, Web APIs and tags) to infer topic probability distribution of Mashup services, which serves as a basis of Mashup service similarity computation. Finally, K-means and Agnes algorithm are used to perform Mashup service clustering in terms of their similarities. ResultsCompared with other existing Mashup service clustering methods, experimental results show that the proposed method achieves a significant improvement in terms of precision, recall, F-measure, purity and entropy. ConclusionThe results of the proposed method help software developers to improve the quality of Mashup service discovery and Mashup-based software development. In the future, there will be a need to extend the method by considering heterogeneous network information among Mashup, Web APIs, tags, users, and applying it to Mashup discovery for software developers.

Identifying Effective Features and Classifiers for Short Term Rainfall Forecast Using Rough Sets Maximum Frequency Weighted Feature Reduction Technique

Precise rainfall forecasting is a common challenge across the globe in meteorological predictions. As rainfall forecasting involves rather complex dynamic parameters, an increasing demand for novel approaches to improve the forecasting accuracy has heightened. Recently, Rough Set Theory (RST) has attracted a wide variety of scientific applications and is extensively adopted in decision support systems. Although there are several weather prediction techniques in the existing literature, identifying significant input for modelling effective rainfall prediction is not addressed in the present mechanisms. Therefore, this investigation has examined the feasibility of using rough set based feature selection and data mining methods, namely Naive Bayes (NB), Bayesian Logistic Regression (BLR), Multi-Layer Perceptron (MLP), J48, Classification and Regression Tree (CART), Random Forest (RF), and Support Vector Machine (SVM), to forecast rainfall. Feature selection or reduction process is a process of identifying a significant feature subset, in which the generated subset must characterize the information system as a complete feature set. This paper introduces a novel rough set based Maximum Frequency Weighted (MFW) feature reduction technique for finding an effective feature subset for modelling an efficient rainfall forecast system. The experimental analysis and the results indicate substantial improvements of prediction models when trained using the selected feature subset. CART and J48 classifiers have achieved an improved accuracy of 83.42% and 89.72%, respectively. From the experimental study, relative humidity2 (a4) and solar radiation (a6) have been identified as the effective parameters for modelling rainfall prediction. ACM CCS (2012) Classification : Computing methodologies → Machine learning → Machine learning algorithms → Feature selection; Information systems → Information retrieval → Retrieval tasks and goals → Clustering and classification; Applied computing → Operations research → Forecasting *To cite this article: S. Mohankumar and V. Balasubramanian, Identifying Effective Features and Classifiers for Short Term Rainfall Forecast Using Rough Sets Maximum Frequency Weighted Feature Reduction Technique, CIT. Journal of Computing and Information Technology , vol. 24, no. 2, pp. 181–194, 2016.

NET-EFFECT IN THE DIAGNOSIS OF BREAST CANCER USING FEATURE REDUCTION AND CLASSIFICATION

Soft Computing methods provide solutions to biologically inspired problem of medical domain like breast cancer. Neural Networks, Fuzzy Logic and Genetic Algorithms contribute novel algorithms to deal with breast cancer. Breast cancer can be diagnosed using soft computing methods. In this paper, we try to produce effective diagnosis of breast cancer by using feature reduction and classification methods. The net-effect of the classification before and after feature reduction process is stated. The feature reduction method applied is Principal Component Analysis (PCA) and the classification method includes Support Vector Machines (SVM). The result of the proposed method produced better outcome when applied on Wisconsin Breast Cancer Data Set (WBCD).

비제약적 환경에서 얼굴 주요위치 특징 서술자 기반의 얼굴인식

This paper proposes a scalable face recognition method for unconstrained face databases, and shows a simple experimental result. Existing face recognition research usually has focused on improving the recognition rate in a constrained environment where illumination, face alignment, facial expression, and background is controlled. Therefore, it cannot be applied in unconstrained face databases. The proposed system is face feature extraction algorithm for unconstrained face recognition. First of all, we extract the area that represent the important features(landmarks) in the face, like the eyes, nose, and mouth. Each landmark is represented by a high-dimensional LBP(Local Binary Pattern) histogram feature vector. The multi-scale LBP histogram vector corresponding to a single landmark, becomes a low-dimensional face feature vector through the feature reduction process, PCA(Principal Component Analysis) and LDA(Linear Discriminant Analysis). We use the Rank acquisition method and Precision at k(p@k) performance verification method for verifying the face recognition performance of the low-dimensional face feature by the proposed algorithm. To generate the experimental results of face recognition we used the FERET, LFW and PubFig83 database. The face recognition system using the proposed algorithm showed a better classification performance over the existing methods.

Extracting salient features for network intrusion detection using machine learning methods

This work presents a data preprocessing and feature selection framework to support data mining and network security experts in minimal feature set selection of intrusion detection data. This process is supported by detailed visualisation and examination of class distributions. Distribution histograms, scatter plots and information gain are presented as supportive feature reduction tools. The feature reduction process applied is based on decision tree pruning and backward elimination. This paper starts with an analysis of the KDD Cup '99 datasets and their potential for feature reduction. The dataset consists of connection records with 41 features whose relevance for intrusion detection are not clear. All traffic is either classified `normal' or into the four attack types denial-of-service, network probe, remote-to-local or user-to-root. Using our custom feature selection process, we show how we can significantly reduce the number features in the dataset to a few salient features. We conclude by presenting minimal sets with 4--8 salient features for two-class and multi-class categorisation for detecting intrusions, as well as for the detection of individual attack classes; the performance using a static classifier compares favourably to the performance using all features available. The suggested process is of general nature and can be applied to any similar dataset.

A Review of Feature Reduction in Intrusion Detection System Based on Artificial Immune System and Neural Network

Feature reduction plays an important role in intrusion detection system. The large amount of feature in network as well as host data effect the performance of intrusion detection method. Various authors are research proposed a method of intrusion detection based on machine learning approach and neural network approach, but all of these methods lacks in large number of feature attribute in intrusion data. In this paper we discuss its various method of feature reduction using artificial immune system and neural network. Artificial immune system is biological inspired system work as mathematical model for feature reduction process. The neural network well knows optimization technique in other field. In this paper we used neural network as feature reduction process. The feature reduction process reduces feature of intrusion data those are not involved in security threats and attacks such as TCP protocol, UDP protocol and ICMP message protocol. This reduces feature-set of intrusion improve the classification rate of intrusion detection and improve the speed performance of the intrusion detection system. The current research going on fixed and static number of feature reduction, we proposed an automatic and dynamic feature reduction technique using PCNN network.

Fuzzy Weighted Gaussian Mixture Model for Feature Reduction

reduction is one kind of pattern recognition and decision making technique, which can be achieved by using Fuzzy Weighted Gaussian Mixture Model (FWGMM) based on the Gaussian Mixture Model. This model helps to find relevant features by using Fuzzy ordered weighted average, which leads to determine the similarity of the density mixture. The salient feature of this approach is to find the relevant features simultaneously by employing fuzzy weighted approach. By applying Ordered Weighted Average (OWA), the feature weights are calculated and they are ordered using the membership values (oring criterion). Hence the feature weights are used as a regulator to determine the relevant features in feature reduction process. Maximum Ordered Weighted Average Likelihood (MOWAL) Framework adopts the Fuzzy Weighted - Gaussian Mixture Model (FW-GMM) for finding the component, which helps to discriminate the relevance of the features and improve the accuracy of density mixture.

Kernelized partial least squares for feature reduction and classification of gene microarray data.

BackgroundThe primary objectives of this paper are: 1.) to apply Statistical Learning Theory (SLT), specifically Partial Least Squares (PLS) and Kernelized PLS (K-PLS), to the universal "feature-rich/case-poor" (also known as "large p small n", or "high-dimension, low-sample size") microarray problem by eliminating those features (or probes) that do not contribute to the "best" chromosome bio-markers for lung cancer, and 2.) quantitatively measure and verify (by an independent means) the efficacy of this PLS process. A secondary objective is to integrate these significant improvements in diagnostic and prognostic biomedical applications into the clinical research arena. That is, to devise a framework for converting SLT results into direct, useful clinical information for patient care or pharmaceutical research. We, therefore, propose and preliminarily evaluate, a process whereby PLS, K-PLS, and Support Vector Machines (SVM) may be integrated with the accepted and well understood traditional biostatistical "gold standard", Cox Proportional Hazard model and Kaplan-Meier survival analysis methods. Specifically, this new combination will be illustrated with both PLS and Kaplan-Meier followed by PLS and Cox Hazard Ratios (CHR) and can be easily extended for both the K-PLS and SVM paradigms. Finally, these previously described processes are contained in the Fine Feature Selection (FFS) component of our overall feature reduction/evaluation process, which consists of the following components: 1.) coarse feature reduction, 2.) fine feature selection and 3.) classification (as described in this paper) and prediction.ResultsOur results for PLS and K-PLS showed that these techniques, as part of our overall feature reduction process, performed well on noisy microarray data. The best performance was a good 0.794 Area Under a Receiver Operating Characteristic (ROC) Curve (AUC) for classification of recurrence prior to or after 36 months and a strong 0.869 AUC for classification of recurrence prior to or after 60 months. Kaplan-Meier curves for the classification groups were clearly separated, with p-values below 4.5e-12 for both 36 and 60 months. CHRs were also good, with ratios of 2.846341 (36 months) and 3.996732 (60 months).ConclusionsSLT techniques such as PLS and K-PLS can effectively address difficult problems with analyzing biomedical data such as microarrays. The combinations with established biostatistical techniques demonstrated in this paper allow these methods to move from academic research and into clinical practice.

Integration and Reduction of Microarray Gene Expressions Using an Information Theory Approach

The DNA microarray is an important technique that allows researchers to analyze many gene expression data in parallel. Although the data can be more significant if they come out of separate experiments, one of the most challenging phases in the microarray context is the integration of separate expression level datasets that have gathered through different techniques. In this paper, we present a general novel method for the integration of any collected data whose distributions have been linearly transformed. The new method is based on the information theory concepts. More than that, this article presents a new approach for checking of the linearity between two distributions as a validation technique. The validation technique assists in taking the feature reduction process in effect prior to the integration phase. The time complexity of the proposed algorithm is low and the new presented methods show good functionality. The experimental results are presented at the end of the paper.