Generalized Discriminant Analysis Research Articles

Wearing sensor data integration for promoting the performance skills of music in IoT

AbstractThis study integrates multi‐node wearable sensor data to improve music performance skills. A window‐adding method is used during time‐frequency feature extraction. By incorporating kernel functions, we present a generalized discriminant analysis (GDA) method to reduce the high‐dimensional sensor features while retaining performance traits. Experiments demonstrate that the proposed GDA approach achieves higher accuracy (92.71%), precision (90.54%), and recall (88.68%) compared to linear discriminant analysis (82.39% accuracy) and principal component analysis (88.56% accuracy) in classifying motions performed by music performers. The integrated analysis of wearable sensor data facilitates comprehensive feedback to strengthen proficiency across various music performance skills.

Designing a new deep convolutional neural network for content-based image retrieval with relevance feedback

Content-based image retrieval (CBIR) is a hot research topic in computer vision. Relevance Feedback (RF) is a powerful technique that can help to increase quality of the CBIR. In this study, a new method for addressing the problem of the CBIR is proposed. To this end, a novel Convolutional Neural Network (CNN) and a new framework of applying RF have been investigated. Also, in order to obtain more efficient image features and reduce the dimensionality of feature vectors, we applied Generalized Discriminant Analysis (GDA) on the extracted features. The proposed method was tested on three benchmark datasets including the OT, Corel-1000 and Caltech-101. The experimental results illustrate that the proposed method performs better compared to the state-of-the-art in the field. Our approach achieved remarkable performances in image retrieval by demonstrating mean Average Precisions (mAPs) of 96.8%, 95.62% and 98.78% on the OT, Corel-1000 and Caltech-101, respectively.

On ALIF-WT-GDA-based Terahertz spectral method for identification of transgenic cotton seeds

Terahertz (THz) time-domain spectroscopy can interact with organic molecules to produce “resonance” absorption, and it is suitable for qualitative and quantitative analysis on the basis of spectral absorption mechanism. However, there are some problems with it, such as un-obviousness signal peaks and positions, signal oscillations and low recognition accuracy. Therefore, this paper has proposed a method named ALIF-WT-GDA-based THz spectrum for identifying transgenic cotton seeds, which can strengthen and accurately identify features of the THz spectral data. First, this paper has put forward a pretreatment method of THz spectral data based on Adaptive Local Iterative Filtering (ALIF) and Wavelet Transform (WT). Specifically speaking, ALIF is used to decompose THz time-domain spectral signals and extract useful mode components for signal reconstruction in order to improve the signal-to-noise ratio (SNR). Also, WT is employed to perform smoothing filtering of THz signals, which makes THz absorbance curves much smoother and absorption peak positions much more obvious. Then, thanks to the advantages of kernel methods, Generalized Discriminant Analysis (GDA) method of Kernel Fisher Discriminant Analysis (KFDA) is used for dimension reduction and clustering analysis. Finally, the proposed ALIF-WT-GDA method is adopted for pretreatment and clustering of the THz spectral data from four different types of transgenic cotton seeds, with its identification accuracy of 98.33%. The above result has proved that the proposed ALIF-WT-GDA method is effective.

Raman microspectroscopy based TNM staging and grading of breast cancer

The tumor-node-metastasis (TNM) system is the most common way that doctors determine the anatomical extent of cancer on the basis of clinical and pathological criteria. In this study, a spectral histopathological study has been carried out to bridge Raman micro spectroscopy with the breast cancer TNM system. A total of seventy breast tissue samples, including healthy tissue, early, middle, and advanced cancer, were investigated to provide detailed insights into compositional and structural variations that accompany breast malignant evolution. After evaluating the main spectral variations in all tissue types, the generalized discriminant analysis (GDA) pathological diagnostic model was established to discriminate the TNM staging and grading information. Moreover, micro-Raman images were reconstructed by K-means clustering analysis (KCA) for visualizing the lobular acinar in healthy tissue and ductal structures in all early, middle and advanced breast cancer tissue groups. While, univariate imaging techniques were adapted to describe the distribution differences of biochemical components such as tryptophan, β-carotene, proteins, and lipids in the scanned regions. The achieved spectral histopathological results not only established a spectra-structure correlations via tissue biochemical profiles but also provided important data and discriminative model references for in vivo Raman-based breast cancer diagnosis.

A Novel Machine Learning Approach for Detection of Coronary Artery Disease Using Reduced Non-linear and Chaos Features

In this research paper, authors present an automated system in this paper that integrates a ranking technique with Principal Component Analysis (PCA), Generalized Discriminant Analysis (GDA) and a 1-Norm Bidirectional Extreme Learning Machine (1-NBELM) to reliably classify normal and coronary artery disease groups. Twenty chaotic and non-linear attributes were hauling out from the Heart Rate Variability (HRV) signal to detect coronary artery disease groups. The HRV data for this study derived from a typical database of Normal Old (ELY), Young (YNG), and Coronary Artery Disease (CAD) people. Fisher, Wilcoxon and Bhattacharya were used to compute the rankings of attributes. GDA then turned the ranking features into a new feature. The Radial Basis Function (RBF) kernel was used to transfer original features to a high-dimensional feature space in GDA and PCA, and then it was deployed to 1-NBELM, which utilized the sigmoidal or multiquadric non-linear activation. Numerical experiments were performed on the combination of database sets as Young-ELY, Healthy-CAD, and Healthy ELY-CAD subjects. The numerical results show that ROC with GDA and 1-NBELM approach achieved an accuracy of 98.12±0.14, 96.21±0.12 and 99.87±0.28 for Young-CAD, Young-ELY and Healthy ELY-CAD groups with the use of sigmoidal and multiquadric activation function. The Fisher with GDA and 1-NBELM and Bhattacharya with GDA and 1-Norm Extreme Learning Machine (1-NELM) approach achieved an accuracy of 99.98±0.21 for all databases.

A novel active sensing method for loosening detection of globe-cone joint considering assembly uncertainties

Globe-cone joint is widely used in satellite propulsion system and rocket engine. Loosening of globe-cone joints will trigger leakage failure, and even initiate catastrophic consequences. Effective loosening detection of globe-cone joints is a challenging issue due to assembly uncertainties. To address this issue, a novel active sensing method insensitive to assembly uncertainties is proposed. It contains three steps. In the first step, the concept of multiscale effort to compress (METC) is proposed and effective loosening features, i.e., the ETC values under different scales are constructed. Subsequently, the dimensions of these features are reduced by generalized discriminant analysis (GDA) method. Last, an ensemble learning model is built to accurately recognize different loosening states. Experimental results demonstrate that the detection accuracy is as high as 0.9429 using the proposed method. More experiments have been conducted to validate the superiorities of multiscale ETC and GDA for feature extraction and reduction. Furthermore, the advantages of the proposed method in effectively detecting loosening of globe-cone joints and have been verified by comparing with other common loosening detection methods.

A Novel Incipient Fault Diagnosis Method for Analog Circuits Based on GMKL-SVM and Wavelet Fusion Features

To enhance the reliability of analog circuits in complex electrical systems, a novel incipient fault diagnosis method is presented in this article. The wavelet packet feature quantities, which consist of the energy, fluctuation coefficient, skewness, and margin factor, are obtained via multiscale time-frequency analysis with wavelet packet transform (WPT). Then, generalized discriminant analysis (GDA) is employed to realize the fusion of wavelet packet feature quantities because it can handle the data nonlinearity and eliminate redundant information. Furthermore, the generalized multiple kernel learning support vector machine (GMKL-SVM), which has the advantages of a strong generalization ability and high accuracy, is developed to identify the incipient fault classes of analog circuits. Moreover, a new particle swarm intelligent optimization algorithm, the sine cosine algorithm (SCA), is adopted to optimize key parameters of GMKL-SVM because of its high convergence speed and strong global optimization ability. The method is fully evaluated with the Sallen-Key bandpass filter circuit, the four-op-amp biquad high-pass filter circuit, and the leapfrog filter circuit. The experimental results demonstrate that the proposed incipient fault diagnosis method for analog circuits can produce higher diagnosis accuracy than other typical incipient fault diagnosis methods.

Human Retina Based Identification System Using Gabor Filters and GDA Technique

A biometric authentication system provides an automatic person authentication based on some characteristic features possessed by the individual. Among all other biometrics, human retina is a secure and reliable source of person recognition as it is unique, universal, lies at the back of the eyeball and hence it is unforgeable. The process of authentication mainly includes pre-processing, feature extraction and then features matching and classification. Also authentication systems are mainly appointed in verification and identification mode according to the specific application. In this paper, preprocessing and image enhancement stages involve several steps to highlight interesting features in retinal images. The feature extraction stage is accomplished using a bank of Gabor filter with number of orientations and scales. Generalized Discriminant Analysis (GDA) technique has been used to reduce the size of feature vectors and enhance the performance of proposed algorithm. Finally, classification is accomplished using k-nearest neighbor (KNN) classifier to determine the identity of the genuine user or reject the forged one as the proposed method operates in identification mode. The main contribution in this paper is using Generalized Discriminant Analysis (GDA) technique to address ‘curse of dimensionality’ problem. GDA is a novel method used in the area of retina recognition.

Automatic ECG-Based Emotion Recognition in Music Listening

This paper presents an automatic ECG-based emotion recognition algorithm for human emotion recognition. First, we adopt a musical induction method to induce participants’ real emotional states and collect their ECG signals without any deliberate laboratory setting. Afterward, we develop an automatic ECG-based emotion recognition algorithm to recognize human emotions elicited by listening to music. Physiological ECG features extracted from the time-, and frequency-domain, and nonlinear analyses of ECG signals are used to find emotion-relevant features and to correlate them with emotional states. Subsequently, we develop a sequential forward floating selection-kernel-based class separability-based (SFFS-KBCS-based) feature selection algorithm and utilize the generalized discriminant analysis (GDA) to effectively select significant ECG features associated with emotions and to reduce the dimensions of the selected features, respectively. Positive/negative valence, high/low arousal, and four types of emotions (joy, tension, sadness, and peacefulness) are recognized using least squares support vector machine (LS-SVM) recognizers. The results show that the correct classification rates for positive/negative valence, high/low arousal, and four types of emotion classification tasks are 82.78, 72.91, and 61.52 percent, respectively.

Integration of Wavelet and Recurrence Quantification Analysis in Emotion Recognition of Bilinguals

Background: This study offers a robust framework for the classification of autonomic signals into five affective states during the picture viewing. To this end, the following emotion categories studied: five classes of the arousal-valence plane (5C), three classes of arousal (3A), and three categories of valence (3V). For the first time, the linguality information also incorporated into the recognition procedure. Precisely, the main objective of this paper was to present a fundamental approach for evaluating and classifying the emotions of monolingual and bilingual college students. Methods: Utilizing the nonlinear dynamics, the recurrence quantification measures of the wavelet coefficients extracted. To optimize the feature space, different feature selection approaches, including generalized discriminant analysis (GDA), principal component analysis (PCA), kernel PCA, and linear discriminant analysis (LDA), were examined. Finally, considering linguality information, the classification was performed using a probabilistic neural network (PNN). Results: Using LDA and the PNN, the highest recognition rates of 95.51%, 95.7%, and 95.98% were attained for the 5C, 3A, and 3V, respectively. Considering the linguality information, a further improvement of the classification rates accomplished. Conclusion: The proposed methodology can provide a valuable tool for discriminating affective states in practical applications within the area of human-computer interfaces.

CORONARY HEART DISEASE DETECTION USING NONLINEAR FEATURES AND ONLINE SEQUENTIAL EXTREME LEARNING MACHINE

In this paper, we propose an automated approach that combines the generalized discriminant analysis (GDA) as feature reduction scheme with radial basis function (RBF) kernel and the online sequential extreme learning machine (OSELM) having Sigmoid, Hardlim, RBF and Sine activation function as binary classifier for detection of congestive heart failure (CHF) and coronary artery disease (CAD). For this analysis, 13 nonlinear features as Correlation Dimension (CD), Detrended Fluctuation Analysis (DFA) as DFA-[Formula: see text]1 and DFA-[Formula: see text]2, Bubble Entropy (BBEn), Sample Entropy (SampEn), Dispersion Entropy (DISEn), Lempel–Ziv Complexity (LZ), Sinai Entropy (SIEn), Improved Multiscale Permutation Entropy (IMPE), Hurst Exponent (HE), Permutation Entropy (PE), Approximate Entropy (ApEn) and Standard Deviation (SD1/SD2) were extracted from Heart Rate Variability (HRV) signals. For validation of proposed method, HRV data were obtained from standard database of normal sinus rhythm (NSR), CHF and CAD subjects. Numerical experiments were done on the combination of database sets such as NSR-CAD, CHF-CAD and NSR-CHF subjects. The simulation results show a clear difference in combination of database sets by using GDA having RBF, Gaussian kernel function and OSELM binary classifier having Sigmoid, RBF and Sine activation function and achieved an accuracy of 98.17% for NSR-CAD, 100% for NSR-CHF and CAD-CHF subjects.

Combining Multifactorial Assessment Tools and Dimensionality Reduction Analysis for Fall Risk Classification in Community-Dwelling Older Adults

Background and Purpose: Optimal approaches in fall risk assessment involve interdisciplinary collaboration of assessment. This current work aimed at screening the fall risk characteristics from the objective balance and mobility tests between older fallers and nonfallers and further assessing the feasibility of 2 statistical dimensionality reduction models, Linear Discriminant Analysis (LDA) and Generalized Discriminant Analysis (GDA) for discriminating older nonspecific fallers. We hypothesized that the high-dimensionality objective sensor-based parameters, followed by a feature selection and dimensionality reduction process, would be able to discriminate older nonspecific fallers. Methods: Thirty-one community-living older individuals who were older than 60 years (faller: n = 15; nonfaller: n = 16) were recruited. The measurements include gait, balance, and ankle proprioception performances. LDA and GDA were further applied to obtain more discriminative feature space. Receiver-operating characteristic (ROC) curves were constructed to compare the classification quality in all the features. Results: Although some features in single objective measure reached statistical significance, the original features still resulted in high within-class and low between-class variances in the feature space. By further applying LDA and GDA on the original features, the performance of LDA in the feature space was improved. The area under the curve of ROC was GDA dimensionality reduction feature (1), LDA dimensionality reduction feature (0.99), proprioception (0.752), inertial measurement unit (0.745), and center of pressure (0.72), respectively. Conclusions: Experimental results showed the GDA feature has the best classification quality and the additional advantage in combination of interdisciplinary multifactorial fall risk assessment.

An unsupervised feature extraction method based on band correlation clustering for hyperspectral image classification using limited training samples

ABSTRACTDue to the advancement of sensor technology, hyperspectral sensors are now able to record reflectance of the earth’s surface in hundreds of bands. This makes hyperspectral images a rich source for a diverse range of applications, including classification tasks. Based on Hughes phenomenon, classification accuracy decreases as dimensionality increases while a limited number of training samples are within reach. In a hyperspectral image adjacent and non-adjacent bands may have a high correlation; therefore removing this redundancy reduces the amount of data for further analysis including classification. Extraction of a few features which lead to high classification accuracy is a need. In this article, we propose an unsupervised feature extraction method called band correlation clustering (BCC). The proposed method includes three main steps: 1) Calculation of the bands’ correlation coefficient and generating correlation coefficient matrix, 2) Grouping the bands based on correlation coefficient matrix using the k-means clustering algorithm, and 3) Mean calculation of each cluster as a newly extracted feature. The proposed method is evaluated by two different criteria: classification accuracy and time consumption. For performance evaluation, the extracted features are fed to two supervised classifiers of non-parametric SVM and parametric ML. The results are compared with the ones obtained by unsupervised extracted features using principle component analysis (PCA), independent component analysis (ICA), and minimum noise fraction (MNF) and supervised ones using generalized discriminant analysis (GDA) and clustering-based feature extraction (CBFE). Comparison of the results shows promising performance for proposed BCC both in terms of computational costs and classification accuracy.

Detection of coronary artery disease by reduced features and extreme learning machine.

ObjectiveCardiovascular diseases generate the highest mortality in the globe population, mainly due to coronary artery disease (CAD) like arrhythmia, myocardial infarction and heart failure. Therefore, an early identification of CAD and diagnosis is essential. For this, we have proposed a new approach to detect the CAD patients using heart rate variability (HRV) signals. This approach is based on subspaces decomposition of HRV signals using multiscale wavelet packet (MSWP) transform and entropy features extracted from decomposed HRV signals. The detection performance was analyzed using Fisher ranking method, generalized discriminant analysis (GDA) and binary classifier as extreme learning machine (ELM). The ranking strategies designate rank to the available features extracted by entropy methods from decomposed heart rate variability (HRV) signals and organize them according to their clinical importance. The GDA diminishes the dimension of ranked features. In addition, it can enhance the classification accuracy by picking the best discerning of ranked features. The main advantage of ELM is that the hidden layer does not require tuning and it also has a fast rate of detection.MethodologyFor the detection of CAD patients, the HRV data of healthy normal sinus rhythm (NSR) and CAD patients were obtained from a standard database. Self recorded data as normal sinus rhythm (Self_NSR) of healthy subjects were also used in this work. Initially, the HRV time-series was decomposed to 4 levels using MSWP transform. Sixty two features were extracted from decomposed HRV signals by non-linear methods for HRV analysis, fuzzy entropy (FZE) and Kraskov nearest neighbour entropy (K-NNE). Out of sixty-two features, 31 entropy features were extracted by FZE and 31 entropy features were extracted by K-NNE method. These features were selected since every feature has a different physical premise and in this manner concentrates and uses HRV signals information in an assorted technique. Out of 62 features, top ten features were selected, ranked by a ranking method called as Fisher score. The top ten features were applied to the proposed model, GDA with Gaussian or RBF kernal + ELM having hidden node as sigmoid or multiquadric. The GDA method transforms top ten features to only one feature and ELM has been used for classification.ResultsNumerical experimentations were performed on the combination of datasets as NSR-CAD and Self_NSR- CAD subjects. The proposed approach has shown better performance using top ten ranked entropy features. The GDA with RBF kernel + ELM having hidden node as multiquadric method and GDA with Gaussian kernel + ELM having hidden node as sigmoid or multiquadric method achieved an approximate detection accuracy of 100% compared to ELM and linear discriminant analysis (LDA)+ELM for both datasets. The subspaces level-4 and level-3 decomposition of HRV signals by MSWP transform can be used for detection and analysis of CAD patients.

A facial expression recognition system using robust face features from depth videos and deep learning

This work proposes a depth camera-based robust facial expression recognition (FER) system that can be adopted for better human machine interaction. Although video-based facial expression analysis has been focused on by many researchers, there are still various problems to be solved in this regard such as noise due to illumination variations over time. Depth video data in the helps to make an FER system person-independent as pixel values in depth images are distributed based on distances from a depth camera. Besides, depth images should resolve some privacy issues as real identity of a user can be hidden. The accuracy of an FER system is much dependent on the extraction of robust features. Here, we propose a novel method to extract salient features from depth faces that are further combined with deep learning for efficient training and recognition. Eight directional strengths are obtained for each pixel in a depth image where signs of some top strengths are arranged to represent unique as well as robust face features, which can be denoted as Modified Local Directional Patterns (MLDP). The MLDP features are further processed by Generalized Discriminant Analysis (GDA) for better face feature extraction. GDA is an efficient tool that helps distinguishing MLDP features of different facial expressions by clustering the features from the same expression as close as possible and separating the features from different expressions as much as possible in a non-linear space. Then, MLDP-GDA features are applied with Deep Belief Network (DBN) for training different facial expressions. Finally, the trained DBN is used to recognize facial expressions in a depth video for testing. The proposed approach was compared with other traditional approaches in a standalone system where the proposed one showed its superiority by achieving mean recognition rate of 96.25% where the other approaches could make 91.67% at the best. The deep learning-based training and recognition of the facial expression features can also be undertaken with cloud computing to support many users and make the system faster than a standalone system.

Facial Expression Recognition Utilizing Local Direction-Based Robust Features and Deep Belief Network

Emotional health plays very vital role to improve people’s quality of lives, especially for the elderly. Negative emotional states can lead to social or mental health problems. To cope with emotional health problems caused by negative emotions in daily life, we propose efficient facial expression recognition system to contribute in emotional healthcare system. Thus, facial expressions play a key role in our daily communications, and recent years have witnessed a great amount of research works for reliable facial expressions recognition (FER) systems. Therefore, facial expression evaluation or analysis from video information is very challenging and its accuracy depends on the extraction of robust features. In this paper, a unique feature extraction method is presented to extract distinguished features from the human face. For person independent expression recognition, depth video data is used as input to the system where in each frame, pixel intensities are distributed based on the distances to the camera. A novel robust feature extraction process is applied in this work which is named as local directional position pattern (LDPP). In LDPP, after extracting local directional strengths for each pixel such as applied in typical local directional pattern (LDP), top directional strength positions are considered in binary along with their strength sign bits. Considering top directional strength positions with strength signs in LDPP can differentiate edge pixels with bright as well as dark regions on their opposite sides by generating different patterns whereas typical LDP only considers directions representing the top strengths irrespective of their signs as well as position orders (i.e., directions with top strengths represent 1 and rest of them 0), which can generate the same patterns in this regard sometimes. Hence, LDP fails to distinguish edge pixels with opposite bright and dark regions in some cases which can be overcome by LDPP. Moreover, the LDPP capabilities are extended through principal component analysis (PCA) and generalized discriminant analysis (GDA) for better face characteristic illustration in expression. The proposed features are finally applied with deep belief network (DBN) for expression training and recognition.

Human Activity Recognition Using Spatiotemporal 3-D Body Joint Features with Hidden Markov Models

Video-based human-activity recognition has become increasingly popular due to the prominent corresponding applications in a variety of fields such as computer vision, image processing, smart-home healthcare, and human–computer interactions. The essential goals of a video-based activity-recognition system include the provision of behavior-based information to enable functionality that proactively assists a person with his/her tasks. The target of this work is the development of a novel approach for human-activity recognition, whereby human-body-joint features that are extracted from depth videos are used. From silhouette images taken at every depth, the direction and magnitude features are first obtained from each connected body-joint pair so that they can be augmented later with motion direction, as well as with the magnitude features of each joint in the next frame. A generalized discriminant analysis (GDA) is applied to make the spatiotemporal features more robust, followed by the feeding of the time-sequence features into a Hidden Markov Model (HMM) for the training of each activity. Lastly, all of the trained-activity HMMs are used for depth-video activity recognition.

A Local Feature-Based Robust Approach for Facial Expression Recognition from Depth Video

Facial expression recognition (FER) plays a very significant role in computer vision, pattern recognition, and image processing applications such as human computer interaction as it provides sufficient information about emotions of people. For video-based facial expression recognition, depth cameras can be better candidates over RGB cameras as a person’s face cannot be easily recognized from distance-based depth videos hence depth cameras also resolve some privacy issues that can arise using RGB faces. A good FER system is very much reliant on the extraction of robust features as well as recognition engine. In this work, an efficient novel approach is proposed to recognize some facial expressions from time-sequential depth videos. First of all, efficient Local Binary Pattern (LBP) features are obtained from the time-sequential depth faces that are further classified by Generalized Discriminant Analysis (GDA) to make the features more robust and finally, the LBP-GDA features are fed into Hidden Markov Models (HMMs) to train and recognize different facial expressions successfully. The depth information-based proposed facial expression recognition approach is compared to the conventional approaches such as Principal Component Analysis (PCA), Independent Component Analysis (ICA), and Linear Discriminant Analysis (LDA) where the proposed one outperforms others by obtaining better recognition rates.

Two Dimensional Linear Discriminant Analyses for Hyperspectral Data

Most supervised feature extraction methods like linear discriminant analysis (LDA) suffer from the limited number of available training samples. The singularity problem causes LDA to fail in small sample size (SSS) situations. Two dimensional linear discriminant analysis (2DLDA) for feature extraction of hyperspectral images is proposed in this paper which has good efficiency with small training sample size. In this approach, the feature vector of each pixel of hyperspectral image is transformed into a feature matrix. As a result, the data matrices lie in a low-dimensional space. Then, the between-class and within-class scatter matrices are calculated using the matrix form of training samples. The proposed approach has two main advantages: it deals with the SSS problem in hyperspectral data, and also it can extract each number of features (with no limitation) from the original high dimensional data. The proposed method is tested on four widely used hyperspectral datasets. Experimental results confirm that the proposed 2DLDA feature extraction method provides better classification accuracy, with a reasonable computation time, compared to popular supervised feature extraction methods such as generalized discriminant analysis (GDA) and nonparametric weighted feature extraction (NWFE) particularly compared to the 1DLDA in the SSS situation. The experiments show that two dimensional linear discriminant analysis+support vector machine (2DLDA+SVM) is an appropriate choice for feature extraction and classification of hyperspectral images using limited training samples.

Cyber Attack Detection System based on Improved Support Vector Machine

This paper presents a novel cyber attack classification approach using improved Support Vector Machine (iSVM) by modifying Gaussian kernel. The Support Vector Machine (SVM) is based on machine learning technique known to perform well at various pattern recognition tasks; such as image classification, text categorization and handwritten character recognition. The cyber attack detection is basically a pattern classification problem, in which classification of normal pattern is done from the abnormal pattern (attack). Although, traditional SVM is better classifier in terms of fast training, scalable and generalization capability. Performance of traditional SVM is enhanced in this work by modifying Gaussian kernel to enlarge the spatial resolution around the margin by a conformal mapping, so that the separability between attack classes is increased. It is based on the Riemannian geometrical structure induced by the kernel function. In the proposed method, class specific Cyber Attack Detection System which combines feature reduction technique and improved support vector machine classifier. This technique has two phases, in the first phase we reduced the redundant features of the original KDDCUP2009 dataset by Generalized Discriminant Analysis (GDA). In the second phase we used improved Support Vector Machine (iSVM) classifier to classify the reduced dataset obtained from first phase. Result shows that iSVM gives 100% detection accuracy for Normal and Denial of Service (DOS) classes and comparable to false alarm rate, training, and testing times.