Best Set Of Features Research Articles

A novel QIM-DCT based fusion approach for classification of remote sensing images via PSO and SVM models

Fusion of panchromatic and multispectral images has become a research interest for the classification of remote sensing images. The spectral and spatial resolutions of different images give better information with the aid of image classification. However, fusing pixels for various satellite images is difficult due to the nature of original image consists of complex information. Similarly, most of the existing fusion algorithms implement a unified processing over the whole part of the image, thereby leaving certain important needs out of consideration. The main aim of our proposed approach is to fuse the images by gathering all important information from multiple images with minimum errors. In this paper, we propose a novel quantization index modulation with discrete contourlet transform-based fusion approach for classification of remote sensing images (LISS IV sensor). In order to improve the image fusion performance, we eliminate certain noises (salt, pepper, and Gaussian) using Bayesian filter with Adaptive Type-2 Fuzzy System. After image fusion, we make image classification by two steps of processes including deep multi-feature extraction and feature selection. Multiple features such as spectral, shape, global and local features are extracted using Affine Transformation (0??, 90??, 180??, and 270??), and then the best set of features are chosen by mutual information and maximal information coefficients. Finally, the image is classified into seven classes using PSO and SVM namely Urban, Vegetation, Wetland, Tank, Water Area, Bare Land, and Roadways. MATLAB R2017b has been used for evaluation of the LISS IV images. Experimental results revealed that our proposed approach is very effective in terms of their classification accuracy.

Low-complexity disordered speech quality estimation

Tracheoesophageal (TE) speech is generated by patients who have undergone a total laryngectomy where the larynx (voice box) is removed and replaced by a tracheoesophageal puncture. This work presents a novel low complexity algorithm to estimate the degree of severity of disordered TE speech. The proposed algorithm has two output scores which are computed from 20 ms voiced frames of the speech signal. An 18th order Linear Prediction (LP) analysis is performed on each voiced frame of the speech signal. The first output score uses features derived from high order statistics (mean, variance, skewness and kurtosis) which are calculated from the LP coefficients, the cepstral coefficients and the LP residual signal. These high order statistics (HOS) along with the pitch value are averaged over all voiced frames yielding a total of 14 HOS quality features. The second output score is derived from features derived from the estimated vocal tract model parameters (cross-sectional tubes areas). Statistical vocal tract parameters (VTPs) across all voiced speech frames were used as speech quality features. Forward stepwise regression as well as K-fold cross validation are then used to select the best sets of features to be fed to the regression models. The results show high correlations with subjective scores for several regression techniques that can provide a correlation up to 0.91 when VTP-Gaussian model is used.

Efficient feature selection techniques for sentiment analysis

Sentiment analysis is a domain of study that focuses on identifying and classifying the ideas expressed in the form of text into positive, negative and neutral polarities. Feature selection is a crucial process in machine learning. In this paper, we aim to study the performance of different feature selection techniques for sentiment analysis. Term Frequency Inverse Document Frequency (TF-IDF) is used as the feature extraction technique for creating feature vocabulary. Various Feature Selection (FS) techniques are experimented to select the best set of features from feature vocabulary. The selected features are trained using different machine learning classifiers Logistic Regression (LR), Support Vector Machines (SVM), Decision Tree (DT) and Naive Bayes (NB). Ensemble techniques Bagging and Random Subspace are applied on classifiers to enhance the performance on sentiment analysis. We show that, when the best FS techniques are trained using ensemble methods achieve remarkable results on sentiment analysis. We also compare the performance of FS methods trained using Bagging, Random Subspace with varied neural network architectures. We show that FS techniques trained using ensemble classifiers outperform neural networks requiring significantly less training time and parameters thereby eliminating the need for extensive hyper-parameter tuning.

Extracting Novel Features for Skin Burn Image Classification

In this paper, the objective is to propose a set of novel features for the classification of different burn depths by using an image mining approach. Both colour and texture features were studied on skin burn dataset comprising skin burn images categorized into three burn depths by the burn specialist. The performance of the proposed feature set was evaluated using linear SVM on 10-fold cross validation method. The empirical results showed that the six proposed novel features, when used together with the common image features, was the best set of features that was able to classify most of the burn depths in terms of accuracy, precision and recall measures with the values of 96.8750%, 96.9697% and 96.6667% respectively. Automated classification of skin burn depths is essential because the initial burn treatment provided to patients are usually based on the first evaluation of the skin burn injuries by determining the burn depths. However, the burn specialist may not always be available at the accident site. In conclusion, the features extracted that represent the burn characteristics specifically in terms of colour and texture were able to effectively characterise the depth of burns in accordance to burn depth classification.

Feature Selection for Phishing Detection with Machine Learning

In the last 10 years machine learning has been widely used to combat phishing attacks. The most common approach was to build a classification model that would be able to detect whether or not a given URL or website is a phishing attack. In order to effectively detect a phishing page with machine learning we must find an effective method to represent websites (both phish and benign) as features which can be fed into a machine learning model. One of the challenges faced by these approaches was to find a good set of features to represent the phishing and benign sites. Within the last 10 years hundreds of different features had been proposed and used to great success [6] [7] [9]. However, due to the curse of dimensionality, use of all available features will exponentially increase the sparsity of the dataset, lowering the odds of successful classification. In this work we extract 31 features that had been commonly used in the literature and perform an in depth feature ranking analysis in order to find the most effective features for phishing detection. Using both filter and wrapping methods we were able to find 23 effective features for phishing detection. The F1-score for all 31 features was 0.88 and time taken to train the multilevel perceptron model was 45.49 seconds and the size of the data used is 100k. Using these 23 features we were able to train a model that has 0.99 F1-score and which was comparable with all previous work while reducing the overall dimensionality of the data and time taken to train the model was 43.71 seconds.

Gender Profiling From a Single Snapshot of Apps Installed on a Smartphone: An Empirical Study

The integration of the fifth generation (5G) networks and artificial intelligence (AI) benefits to create a more holistic and better connected ecosystem for industries. User profiling has become an important issue for industries to improve company profit. In the 5G era, smartphone applications have become an indispensable part in our everyday lives. Users determine what apps to install based on their personal needs, interests, and tastes, which is likely shaped by their genders-the behavioral, cultural, or psychological traits typically associated with their sex. It is possible to profile users' gender based simply on a single snapshot of apps installed on their smartphones. With this inference based on easy to access data, we can make smartphone systems more user-friendly, and provide better personalized products and services. In this article, we explore such possibilities through an empirical study on a large-scale dataset of installed app lists from 15 000 Android users. More specifically, we investigate the following research questions: 1) What differences between females and males can be explored from installed app lists? 2) Can user gender be reliably inferred from a snapshot of apps installed? Which snapshot feature(s) are the most predictive? What is the best combination of features for building the gender prediction model? 3) What are the limitations of a gender prediction model based solely on a snapshot of apps installed on a smartphone? We find significant gender differences in app type, function, and icon design. We then extract the corresponding features from a snapshot of apps installed to infer the gender of each user. We assess the gender predictive ability of individual features and combinations of different features. We achieve an accuracy of 76.62% and area under the curve of 84.23% with the best set of features, outperforming the existing work by around 5% and 10%, respectively. Finally, we perform an error analysis on misclassified users and discussed the implications and limitations of this article.

Estimation of maintainability parameters for object-oriented software using hybrid neural network and class level metrics

The various software metrics proposed in the literature can be used to evaluate the quality of software systems written in object-oriented manner. These metrics are broadly categorized into two subcategories i.e., system level software metrics and class level software metrics. In this work, ten different types of class level metrics are considered as an input to develop one model for predicting software maintainability of object-oriented software system. These models are developed using three types of neural networks, i.e., artificial neural network, radial basis function network, and functional link artificial neural network. In this study, a hybrid algorithm based on genetic algorithm (GA) with gradient descent algorithm has been proposed to find optimal weights of these neural networks. Since accuracy of the prediction model is highly dependent on the class level metrics, they are considered as input of the models. So, five different feature selection techniques are used in this study to identify the best set of features with an objective to improve the accuracy of software maintainability prediction model. The effectiveness of these models are evaluated using four evaluation metrics, i.e., MAE, MMRE, RMSE, and SEM. In this work, parallel computing concept has been also considered with an objective to reduce the model training time. The results show that the model developed using the proposed hybrid algorithm based on GA with gradient descent algorithm give better results as compared to the work presented by other authors in literature. The results also show that feature selection techniques obtain better results for predicting maintainability as compared to all metrics. The experimental results show that parallel computing is beneficial in reducing the model training time.

An hybrid feature space from texture information and transfer learning for glaucoma classification

Glaucoma is a progressive eye disease due to the increase in intraocular pressure. Accurate early detection may prevent vision loss. Most algorithms in the literature are not feasible for use in screening programs since they are not able to handle a wide diversity of images. We conducted an extensive study to determine the best set of features for image representation. Our feature extraction methodology included the following descriptors: LBP, GLCM, HOG, Tamura, GLRLM, morphology, and seven CNN architectures, that results in 30.682 features. Then, we used the gain ratio to order the features by importance and select the best set for glaucoma classification. Our tests were performed using 1675 images of DRISHTI, RIM-ONE, HRF, JSIEC, and ACRIMA databases. We concluded that a combination of the GLCM and pre-trained CNN’s has the potential to be used in a computer aid system for glaucoma detection. Our approach achieved an accuracy of 93.61%.

A performance based feature selection technique for subject independent MI based BCI.

Significant research has been conducted in the field of brain computer interface (BCI) algorithm development, however, many of the resulting algorithms are both complex, and specific to a particular user as the most successful methodology can vary between individuals and sessions. The objective of this study was to develop a simple yet effective method of feature selection to improve the accuracy of a subject independent BCI algorithm and streamline the process of BCI algorithm development. Over the past several years, several high precision features have been suggested by researchers to classify different motor imagery tasks. This research applies fourteen of these features as a feature pool that can be used as a reference for future researchers. Additionally, we look for the most efficient feature or feature set with four different classifiers that best differentiates several motor imagery tasks. In this work we have successfully employed a feature fusion method to obtain the best sub-set of features. We have proposed a novel computer aided feature selection method to determine the best set of features for distinguishing between motor imagery tasks in lieu of the manual feature selection that has been performed in past studies. The features selected by this method were then fed into a Linear Discriminant Analysis, K-nearest neighbor, decision tree, or support vector machine classifier for classification to determine the overall performance. The methods used were a novel performance based additive feature fusion algorithm working in conjunction with machine learning in order to classify the motor imagery signals into particular states. The data used for this study was collected from BCI competition III dataset IVa. The result of this algorithm was a classification accuracy of 99% for a subject independent algorithm with less computation cost compared to traditional methods, in addition to multiple feature/classifier combinations that outperform current subject independent methods. The conclusion of this study and its significance is that it developed a viable methodology for simple, efficient feature selection and BCI algorithm development, which leads to an overall increase in algorithm classification accuracy.

Investigating smartphone user differences in their application usage behaviors: an empirical study

Smartphone applications (Abbr. apps) have become an indispensable part in our everyday lives. Users determine what apps to use depending on their personal needs and interests. Users with different attributes may have different needs, making it natural for their app usage behaviors to be different. The differences in app usage behaviors among users make it possible to infer their attributes. Knowing such differences could help improve mobile user experiences by enhancing smart services and devices. In this paper, we present an empirical study of investigating smartphone user differences on a large-scale dataset of recently used app lists from 106,672 Android users from China. We first investigate the user differences in app usage behaviors with respect to their attributes (gender, age, and income level). We find significant differences in app usage frequency, app usage with time context and functions. We then extract corresponding features from app usage records to infer the attributes of each user, and investigate the predictive ability of individual features and combinations of different individual features. We achieve the accuracy of 83.29% for gender, 69.94% for age (four age ranges) and 71.43% for income level (three income levels) with the best set of features, respectively. Finally, we discuss the implications of our findings and the limitations of this work.

Fractal dimension undirected correlation graph-based support vector machine model for identification of focal and non-focal electroencephalography signals

Recognition of focal (FC) and non-focal (NFC) Electroencephalography (EEG) signals is crucial for clinical diagnosis used to localise and aid in medical treatment of the affected region in the human brain. Developing an artificial intelligence system that can adequately identify these affected regions can support the clinical diagnosis of brain disease. In this study, we develop a new model called a fractal dimension (FD) of the undirected graph (NG) based on a sine cosine driven support vector machine (FD-NG model utilising the SCA-SVM) algorithm for identifying the focal and non-focal EEG signals. Each EEG signal is partitioned into its respective segments and each segment is divided into clusters using a sliding window technique. To reduce the dimensionality of each cluster, a set of best features is extracted. Three types of input features are considered: linear features (LF), statistical features (SF), and features based on time domain (TD). These are investigated and extracted from each cluster. As a result, each EEG signal is represented by a series of reduced segments and is then forwarded to the proposed FD-NG based SCA-SVM model. The model considers each segment as a node and a link is built between each pair of nodes based on their degree of similarity. The FD of graphs are used as inputs to the SCA-SVM model to classify the EEG signal into FC and NFC components. The obtained results, which also demonstrates the practicality of the approach, confirm that the proposed model surpasses the performance of existing state-of- the-art techniques.

A Methodological Approach for Early Melanoma Detection Using Smartphone Captured Images

Researches on applications of mobile devices bring wide variety of uses in healthcare. One such work focus on detection of malignant melanoma using mobile image analysis. Dermoscopy is one of a current use, but need a special expertise for the detection of cancer melanoma. The image taken using smartphone is used for this purpose. It mainly focus on localization of the skin lesion by combining fast skin detection and fusion of two fast segmentation results. This also introduces some set of image features and to capture color variation and border irregularity which are useful for smartphone-captured images. It propose a new feature selection criterion to select a small set of good features used in the final lightweight system. The method introduces a new module for the detection of distorted images such as motion blur and alert users in such situations. The blurred image undergo deblurring to detect the correct result. The result of this application will identify whether the image is malignant melanoma or benign with their intensity value from smartphone captured images used.

GENDER CLASSIFICATION ON SKELETAL REMAINS: EFFICIENCY OF METAHEURISTIC ALGORITHM METHOD AND OPTIMIZED BACK PROPAGATION NEURAL NETWORK

In forensic anthropology, gender classification is one of the crucial steps involved in developing the biological profiles of skeleton remains. There are several different parts of skeleton remains and every part contains several features. However, not all features can contribute to gender classification in forensic anthropology. Besides that, another limitation that exists in previous researches is the absence of parameter optimization for the classifier. Thus, this paper proposed metaheuristic algorithms such as Particle Swarm Optimization, Ant Colony Algorithm and Harmony Search Algorithm based feature selection to identify the most significant features of skeleton remains. Once the set of significant features was obtained, the learning rate and momentum of Back Propagation Neural Network (BPNN) were optimized. This was to obtain a good combination of parameters in order to produce a better gender classification. This study used 1,538 data samples from Goldman Osteometric Dataset which consisted of femur, humerus and tibia parts. Based on the feature selection results, the Optimized BPNN outperformed other methods for all datasets. The Ant Colony Algorithm-Optimized Back Propagation Neural Network produced the highest accuracy for all parts of the skeleton where for femur was 89.44%, the humerus with 88.97% and tibia with 87.52% accuracy. Hence, it can be concluded that optimized parameter is capable of providing a better gender classification performance with the best set of features. Due to good gender classification techniques, the implication of this study is evident in the area of forensic anthropology where the process of developing a biological profile can be shortened which in turn enhances the productivity of anthropologists.

WOA + BRNN: An imbalanced big data classification framework using Whale optimization and deep neural network

Nowadays, big data plays a substantial part in information knowledge analysis, manipulation, and forecasting. Analyzing and extracting knowledge from such big datasets are a very challenging task due to the imbalance of data distribution, which could lead to a biased classification results and wrong decisions. The standard classifiers are not capable of handling such datasets. Hence, a new technique for dealing with such datasets is required. This paper proposes a novel classification framework for big data that consists of three developed phases. The first phase is the feature selection phase, which uses the Whale optimization algorithm (WOA) for finding the best set of features. The second phase is the preprocessing phase, which uses the SMOTE algorithm and the LSH-SMOTE algorithm for solving the class imbalance problem. Lastly, the third phase is WOA + BRNN algorithm, which is using the Whale optimization algorithm for training a deep learning approach called bidirectional recurrent neural network for the first time. Our proposed algorithm WOA-BRNN has been tested against nine highly imbalanced datasets one of them is big dataset in terms of area under curve (AUC) against four of the most common use machine learning algorithms (Naive Bayes, AdaBoostM1, decision table, random tree), in addition to GWO-MLP (training multilayer perceptron using Gray Wolf Optimizer), then we test our algorithm over four well-known datasets against GWO-MLP and particle swarm optimization (PSO-MLP), genetic algorithm (GA-MLP), ant colony optimization (ACO-MLP), evolution strategy (ES-MLP), and population-based incremental learning (PBIL-MLP) in terms of classification accuracy. Experimental results proved that our proposed algorithm WOA + BRNN has achieved promising accuracy and high local optima avoidance, and outperformed four of the most common use machine learning algorithms, and GWO-MLP in terms of AUC.

Construction of Complex Features for Computational Predicting ncRNA-Protein Interaction.

Non-coding RNA (ncRNA) plays important roles in many critical regulation processes. Many ncRNAs perform their regulatory functions by the form of RNA-protein complexes. Therefore, identifying the interaction between ncRNA and protein is fundamental to understand functions of ncRNA. Under pressures from expensive cost of experimental techniques, developing an accuracy computational predictive model has become an indispensable way to identify ncRNA-protein interaction. A powerful predicting model of ncRNA-protein interaction needs a good feature set of characterizing the interaction. In this paper, a novel method is put forward to generate complex features for characterizing ncRNA-protein interaction (named CFRP). To obtain a comprehensive description of ncRNA-protein interaction, complex features are generated by non-linear transformations from the traditional k-mer features of ncRNA and protein sequences. To further reduce the dimensions of complex features, a group of discriminative features are selected by random forest. To validate the performances of the proposed method, a series of experiments are carried on several widely-used public datasets. Compared with the traditional k-mer features, the CFRP complex features can boost the performances of ncRNA-protein interaction prediction model. Meanwhile, the CFRP-based prediction model is compared with several state-of-the-art methods, and the results show that the proposed method achieves better performances than the others in term of the evaluation metrics. In conclusion, the complex features generated by CFRP are beneficial for building a powerful predicting model of ncRNA-protein interaction.

Examining multiple feature evaluation and classification methods for improving the diagnosis of Parkinson’s disease

An accurate diagnosis of Parkinson’s disease by specialists involves many neurological, psychological and physical examinations. The specialists investigate a number of symptoms and signs when examining the nervous system conditions of a person. The diagnosis involves reviewing the medical history and genetic factor of the person. The recent diagnosis methodology to Parkinson’s disease relies on voice disorders analysis. This methodology entails extracting feature sets of a recorded person’s voice then utilizing a machine learning technique to identify the healthy and Parkinson’s cases from the voice. This paper attempts to improve the diagnoses of Parkinson’s disease by testing multiple feature evaluation and classification machine learning methods based on the voice disorders analysis. The aim of this paper is to find the optimal solution to the problem by (i) proposing a new Multiple Feature Evaluation Approach (MFEA) of a multi-agent system (ii) implementing five independent classification schemas which are Decision Tree, Naïve Bayes, Neural Network, Random Forests, and Support Vector Machine on the Parkinson’s diagnosis before and after applying the MFEA, and (iii) evaluating the diagnosis accuracy of the results. The methodology of the tests encompasses 10-fold cross-validation to evaluate the learning of methods and track variation in their performance. The test results show that the MFEA of the multi-agent system finds the best set of features and improves the performance of the classifiers. The average rate of improvement in the diagnostic accuracy of the classifiers are Decision Tree 10.51%, Naïve Bayes 15.22%, Neural Network 9.19%, Random Forests 12.75%, and Support Vector Machine 9.13%. These results show that the MFEA makes a significant improvement to the classifiers’ diagnosis results.

A bearing vibration data analysis based on spectral kurtosis and ConvNet

Today’s modern industry has accepted condition monitoring based on intelligent fault diagnosis of rotating machinery systems to provide precision and sustainability. The conventional signal processing methods are less productive due to the involvement of various noises from different sources in the vibration signal, and therefore, recently fault diagnosis approaches use the artificial intelligent techniques along with the signal processing methods. Thus, motivated by the kurtogram and convolutional neural network (CNN), a novel efficient method for the fault classification of rotating machines is proposed in this paper. Kurtogram is a measure of dispersion for time–frequency energy density which provides additional frequency contents information and as an effect represents a pattern of each fault uniquely. Thus, it inspires to utilize the kurtogram as an input feature vector, and it helps in reducing the task of identifying dominant features to represent the different faults. Hence, this 2D distinct feature vector presented to CNN for fault classification. The various levels of kurtogram are examined by tuning different hyperparameters of CNN to achieve a good feature set for decent performance. The experimental results demonstrate that the proposed method effectively classifies the bearing faults under different operating conditions in comparison to other methods.

"A Neural Network based Cardiac Arrhythmia Diagnosis system from Dynamic Features of Electrocardiogram Signal"

Objectives: Cardiac arrhythmia is a type of disorder where the heartbeat is irregular, too slow, or too fast. As a result of heart diseases, there is an increase in death yearly. The early detection of cardiac diseases is important for preventing the deaths due to the cardiac diseases. Methods: The Electrocardiogram (ECG) is used to record the electrical activity of the heart for physician to diagnose the heart diseases. In this study, we propose a cardiac diagnosis system for diagnosing cardiac arrhythmia disease. It will be most helpful for the patients who undergone a heart surgery for continuous monitoring of post-surgical status. Findings: The major objective of this paper is to implement an effective algorithm to discriminate between the normal and diseased persons. The Monitoring process includes the following tasks, such as preprocessing and feature extraction by Pan Tompkin’s algorithm and the features are classified using neural network and support vector machine. The performance of the classifiers was evaluated using the parameters such as sensitivity, specificity and accuracy. The Accuracy of the neural network algorithm is 88.54% and the accuracy of the Support Vector Machine is 84.37%. Application/Improvements: The Neural network classifier shows better performance compared to support vector machine. In future the classifier is trained using the best set of features using feature selection techniques. Keywords: Cardiac Arrhythmia, Electrocardiogram, Neural Network, Pan Tompkins QRS Detection, Support Vector Machine

Improved salp swarm algorithm based on particle swarm optimization for feature selection

Feature selection (FS) is a machine learning process commonly used to reduce the high dimensionality problems of datasets. This task permits to extract the most representative information of high sized pools of data, reducing the computational effort in other tasks as classification. This article presents a hybrid optimization method for the FS problem; it combines the slap swarm algorithm (SSA) with the particle swarm optimization. The hybridization between both approaches creates an algorithm called SSAPSO, in which the efficacy of the exploration and the exploitation steps is improved. To verify the performance of the proposed algorithm, it is tested over two experimental series, in the first one, it is compared with other similar approaches using benchmark functions. Meanwhile, in the second set of experiments, the SSAPSO is used to determine the best set of features using different UCI datasets. Where the redundant or the confusing features are removed from the original dataset while keeping or yielding a better accuracy. The experimental results provide the evidence of the enhancement in the SSAPSO regarding the performance and the accuracy without affecting the computational effort.

Diagnosis of shockable rhythms for automated external defibrillators using a reliable support vector machine classifier

Abstract Sudden cardiac arrest is mainly caused by ventricular fibrillation and ventricular tachycardia, which are known as shockable rhythms. In this paper, a detection algorithm of shockable rhythms including support vector machine (SVM) model uses the public electrocardiogram (ECG) databases for training and testing. The databases are the Creighton University Ventricular Tachyarrhythmia Database (CUDB) and the MIT-BIH Malignant Ventricular Arrhythmia Database (VFDB). At first, to compose a set of good features, we extend a well-known set of 2 good features such as Count2 and VF-filter Leakage Measure (Lk). We supplemented 5 more good features, selected based on a binary genetic algorithm-based feature selection, among 11 new input candidate features. All the combinations of 7 good features are estimated for their performance on the training and the testing data using the SVM models to identify 6 combinations of the final feature pool. 5-Folds cross validation is then implemented carefully to validate the performance of the SVM classifier using final feature pool on separated and entire 5s-segment databases. The final combination of 4 features, which includes Count2, Lk, Threshold Crossing Interval (TCI), and Centroid Frequency (CF), is addressed by the highest validation performance of the corresponding SVM model. The Count2 shows the proportion of the signal, which is above the mean absolute values of the output of an integer coefficient recursive bandpass filter computed for every 1 s time interval. The Lk represents the output of a narrow bandstop filter, which is applied to the ECG signal with the central frequency being the mean signal frequency. The TCI shows the average time between the fixed thresholds, which are computed for every 1s-segment using the pulses converted from the ECG signal. The CF is the frequency, which bisects vertically the area under the power spectrum. For the proposed algorithm, the average accuracy of 95.9%, sensitivity of 91.7%, and specificity of 96.8% are archived on the evaluation data of the entire database. Comparing to the performance of the SVM model using a combination of the Count2 and the Lk, we report a significant improvement for the accuracy of the SVM model using the final feature combination in average, i.e. 2.6%, 22.4%, and 2.7% on the evaluation data of the entire database, the CUDB, and the VFDB, respectively. Furthermore, existence of ventricular ectopic beats in the input data shows a negligible influence on the final performance of classification.