Higher Classification Accuracy Rate Research Articles

Importance of Graphical User Interface in the design of P300 based Brain–Computer Interface systems

Objectives: Develop an effective and intuitive Graphical User Interface (GUI) for a Brain–Computer Interface (BCI) system, that achieves high classification accuracy and Information Transfer Rates (ITRs), while using a simple classification technique. Objectives also include the development of an output device, that is capable of real time execution of the selected commands. Methods:A region based T9 BCI system with familiar face presentation cues capable of eliciting strong P300 responses was developed. Electroencephalogram (EEG) signals were collected from the Oz, POz, CPz and Cz electrode locations on the scalp and subsequently filtered, averaged and used to extract two features. These feature sets were classified using the Nearest Neighbour Approach (NNA). To complement the developed BCI system, a ‘drone prototype’ capable of simulating six different movements, each over a range of eight distinct selectable distances, was also developed. This was achieved through the construction of a body with 4 movable legs, capable of tilting the main body forward, backward, up and down, as well as a pointer capable of turning left and right. Results:From ten participants, with normal or corrected to normal vision, an average accuracy of 91.3 ± 4.8% and an ITR of 2.2 ± 1.1 commands/minute (12.2 ± 6.0 bits/minute) was achieved. Conclusion:The proposed system was shown to elicit strong P300 responses. When compared to similar P300 BCI systems, which utilise a variety of more complex classifiers, competitive accuracy and ITR results were achieved, implying the superiority of the proposed GUI. Significance: This study supports the hypothesis that more research, time and care should be taken when developing GUIs for BCI systems.

3D Input Convolutional Neural Networks for P300 Signal Detection

P300 signal is an endogenous event related potential component. It is mostly elicited from the frontal to parietal brain lobes. Electroencephalography is used for acquiring P300 signal from scalp. P300 signal is used for brain-computer interface systems. P300 based brain-computer interface systems are preferable since they have high overall performance. The most significant overall performance indicator is information transfer rate for P300 based brain-computer interface systems. P300 signal detection accuracy and P300 detection time are using for information transfer rate calculation. Hence, P300 signal classification accuracy is important for getting higher information transfer rate. In this study, it is aimed to investigate P300 detection model for higher classification accuracy. Thus, it is proposed 3-dimensional input convolutional neural network model for P300 detection. Moreover, the proposed model was applied with region based P300 speller which constituted audio and visual stimuli. In experiments, the participants were asked to spell desired words in two sessions which were offline and online session. Linear support vector machine, stepwise linear discriminant analysis, 2-dimensional input convolutional neural network, and the proposed method were compared in both online and offline sessions. It is reached highest average classification accuracy rate with the proposed method in both sessions. According to the online session result, average classification accuracy was 94.22% in 3-dimensional input convolutional neural network model. Furthermore, average information transfer rate was 5.53 bit/min in 3-dimensional input convolutional neural network model. We have also applied methods on BCI competition III-dataset II for 2 participants “A” and “B” for evaluating performance of algorithms. The proposed method had higher classification accuracy rate than linear support vector machine, stepwise linear discriminant analysis, 2-dimensional input convolutional neural network, and multi-classifier convolutional neural network which was used in other study on same dataset.

Optimized SVM-Driven Multi-Class Approach by Improved ABC to Estimating Ship Systems State

In the intelligent ship field, with the upgrading of ship maintenance mode, the human-centered system maintenance will be gradually replaced by the artificial intelligence decision methods. To improve the training speed and testing accuracy of the state estimation model, an optimized Support Vector Machine (SVM) driven approach by Improved Artificial Bee Colony (IABC) was proposed to solve the global parameters optimization problem. First, the IABC method was achieved from three aspects: nectar source initializing, employed bee global neighborhood searching, and scouts mutation neighborhood searching. Second, the multi-class SVM with one-against-one classifiers was selected, and the best global parameters were achieved by the IABC. Third, the optimized SVM model was adopted in the testing to verify the effectiveness of state estimation. Finally, the elaborated methodology was applied to two actual ship systems to get the analysis results. The effectiveness was verified by using two examples. The results show the following: the IABC optimized SVM can obtain the global optimal parameters at a faster speed than the traditional ABC optimized method; the IABC optimized method can help the training start with better initial parameters, and get a higher classification accuracy rate than the traditional ABC optimized method. Based on the comparative analysis results, the IABC optimized SVM shows an obvious advantage of parameter optimization in the training process, and it can also significantly improve the model training efficiency and achieve a higher state estimation accuracy. The optimized SVM by IABC is an effective state estimation method in ship systems.

Intelligent Waste Management System for Recycling and Resource Optimization

This paper proposes a deep learning-based intelligent waste management system that can accurately classify waste types and optimize waste disposal processes. The proposed system utilizes a convolutional model to concisely identify the waste type from images captured by a camera system. Our system uses intelligent data augmentation to perform large datasets of waste item images and achieves a high classification accuracy rate. The waste types are classified into several categories, including glass, cardboard, metal, plastic, paper, and trash. Experimental results show that our system achieves high accuracy rates in waste classification and improves waste disposal efficiency compared to traditional waste management systems. Our system has the potential to significantly reduce the negative impact of waste on the environment and to promote sustainable waste management practices.

A comparison of machine learning techniques for classification of HIV patients with antiretroviral therapy-induced mitochondrial toxicity from those without mitochondrial toxicity

BackgroundAntiretroviral therapy (ART) has significantly reduced HIV-related morbidity and mortality. However, therapeutic benefit of ART is often limited by delayed drug-associated toxicity. Nucleoside reverse transcriptase inhibitors (NRTIs) are the backbone of ART regimens. NRTIs compete with endogenous deoxyribonucleotide triphosphates (dNTPs) in incorporation into elongating DNA chain resulting in their cytotoxic or antiviral effect. Thus, the efficacy of NRTIs could be affected by direct competition with endogenous dNTPs and/or feedback inhibition of their metabolic enzymes. In this paper, we assessed whether the levels of ribonucleotides (RN) and dNTP pool sizes can be used as biomarkers in distinguishing between HIV-infected patients with ART-induced mitochondrial toxicity and HIV-infected patients without toxicity.MethodsWe used data collected through a case-control study from 50 subjects. Cases were defined as HIV-infected individuals with clinical and/or laboratory evidence of mitochondrial toxicity. Each case was age, gender, and race matched with an HIV-positive without evidence of toxicity. We used a range of machine learning procedures to distinguish between patients with and without toxicity. Using resampling methods like Monte Carlo k-fold cross validation, we compared the accuracy of several machine learning algorithms applied to our data. We used the algorithm with highest classification accuracy rate in evaluating the diagnostic performance of 12 RN and 14 dNTP pool sizes as biomarkers of mitochondrial toxicity.ResultsWe used eight classification algorithms to assess the diagnostic performance of RN and dNTP pool sizes distinguishing HIV patients with and without NRTI-associated mitochondrial toxicity. The algorithms resulted in cross-validated classification rates of 0.65–0.76 for dNTP and 0.72–0.83 for RN, following reduction of the dimensionality of the input data. The reduction of input variables improved the classification performance of the algorithms, with the most pronounced improvement for RN. Complex tree-based methods worked the best for both the deoxyribose dataset (Random Forest) and the ribose dataset (Classification Tree and AdaBoost), but it is worth noting that simple methods such as Linear Discriminant Analysis and Logistic Regression were very competitive in terms of classification performance.ConclusionsOur finding of changes in RN and dNTP pools in participants with mitochondrial toxicity validates the importance of dNTP pools in mitochondrial function. Hence, levels of RN and dNTP pools can be used as biomarkers of ART-induced mitochondrial toxicity.

FusionLearn: a biomarker selection algorithm on cross-platform data.

In high dimensional genetic data analysis, the objective is to select important biomarkers which are involved in some biological processes, such as disease progression, immune response, etc. The experimental data are often collected from different platforms including microarray experiments and proteomic experiments. The conventional single-platform approach lacks the capability to learn from multiple platforms, and the resulted lists of biomarkers vary across different platforms. There is a great need to develop an algorithm which can aggregate information across platforms and provide a consolidated list of biomarkers across different platforms. In this paper, we introduce an R package FusionLearn, which implements a fusion learning algorithm to analyze cross-platform data. The consolidated list of biomarkers is selected by the technique of group penalization. We first apply the algorithm on a collection of breast cancer microarray experiments from the NCBI (National Centre for Biotechnology Information) microarray database and the resulted list of selected genes have higher classification accuracy rate across different datasets than the lists generated from each single dataset. Secondly, we use the software to analyze a combined microarray and proteomic dataset for the study of the growth phase versus the stationary phase in Streptomyces coelicolor. The selected biomarkers demonstrate consistent differential behavior across different platforms. R package: https://cran.r-project.org/package=FusionLearn.

Automatic Face Recognition Based on Sparse Representation and Extended Transfer Learning

Sparse representation has exhibited excellent performance in face recognition. However, this method requires some areas for improvement, especially on insufficient face samples. We aim to design a simple and efficient method to improve sparse representation to solve problems with a small sample size. This paper provides two primary contributions that are very effective in small sample face recognition. First, in order to enhance recognition robustness, we designed an intuitive and mathematically controllable transfer learning method of sparse representation by introducing labeled samples. Second, to obtain high recognition accuracy, we developed a weighted fusion scheme to integrate the sparse representation results generated from original and labeled samples. In the ORL dataset, our algorithm’s highest accuracy rate is 95%. In the FERET dataset, our highest classification accuracy rate is 95%. In the more complex LFW dataset, our highest classification accuracy rate has also reached 83.33%. This shows that our experimental results demonstrate that the proposed method can obtain sufficient performance, whereas the weighted fusion scheme can take advantage of sparse representation on the basis of original and labeled samples. This paper will be very useful for identification based on the Internet-of-Medical-Things.

Towards automated statistical partial discharge source classification using pattern recognition techniques

This study presents a comprehensive review of the automated classification in partial discharge (PD) source identification and probabilistic interpretation of the classification results based on the relationship between the variation of the phase-resolved PD (PRPD) patterns and the source of the PD. The proposed automated classification system consists of modern, high-performance statistical feature extraction methods and classifier algorithms. Their application in online monitoring and recognition of the PD patterns is investigated based on their low-processing time and high-performance evaluation. The application of modern statistical algorithms and pre-processing methods configured in this automated classification system improves the pattern recognition accuracy of the different PD sources that are suitable to be employed in different high-voltage (HV) insulation media. To evaluate the performance of the different combinations of the feature extraction/classier pairs, laboratory setups are designed and built that simulate various types of PDs. The test cells include three sources of PD in , two sources of PD in transformer oil, and corona in the air. Data samples for different classes of PD sources are captured under two levels of voltage and two different levels of noise. The results of this study evaluate the suitability of the proposed classification systems for probabilistic source identification in various insulation media. Furthermore, of importance to the problem of the PD source identification is to assign a ‘degree of membership’ to each PRPD pattern, besides assigning a class label to it. Some of the classifier algorithms studied in this study, such as fuzzy classifiers, are not only able to show high classification accuracy rate, but they also calculate the ‘degree of membership’ of a sample to a class of data. This enables probabilistic interpretation of a new PRPD pattern that is being classified. The determination of the degree of membership for future PRPD samples allows safer decision making based on the risk associated with the different sources of PD in HV apparatus.

SVM-RFE-ED: A Novel SVM-RFE based on Energy Distance for Gene Selection and Cancer Diagnosis

Microarray expression data has been a very active research field and anindispensable tool for cancer diagnosis. The microarray expression datasetcontains thousands of genes and selecting a subset of informative genes is aprimordial preprocessing step for improving the cancer classification.Support Vector Machine Recursive Feature Elimination (SVM-RFE) is one of the popular and effective gene selection approaches.However, SVM-RFE attempts to find the best possible combination forclassification and does not take into account the ability of classseparability for each gene. In this paper, a novel SVM-RFE based on energydistance (ED) and called SVM-RFE-ED is proposed to overcome the limitation ofstandard SVM-RFE. The aims of our study are to achieve a high classificationaccuracy rate and improve the classification model. The experimentation isconducted on five widely used datasets. Experimental results indicate thatthe proposed approach SVM-RFE-ED provides good results and achieve a highclassification accuracy rate using a small number of genes.

Convolutional Shape-Aware Representation for 3D Object Classification

Deep learning has recently emerged as one of the most popular and powerful paradigms for learning tasks. In this paper, we present a deep learning approach to 3D shape classification using convolutional neural networks. The proposed framework takes a multi-stage approach that first represents each 3D shape in the dataset as a 2D image using the bag-of-features model in conjunction with intrinsic spatial pyramid matching that leverages the spatial relationship between features. These 2D images are then fed into a pre-trained convolutional neural network to learn deep convolutional shape-aware descriptors from the penultimate fully-connected layer of the network. Finally, a multiclass support vector machine classifier is trained on the deep descriptors, and the classification accuracy is subsequently computed. The effectiveness of our approach is demonstrated on three standard 3D shape benchmarks, yielding higher classification accuracy rates compared to existing methods.

Enhancing Performance and Bit Rates in a Brain-Computer Interface System With Phase-to-Amplitude Cross-Frequency Coupling: Evidences From Traditional c-VEP, Fast c-VEP, and SSVEP Designs.

A brain–computer interface (BCI) is a channel of communication that transforms brain activity into specific commands for manipulating a personal computer or other home or electrical devices. In other words, a BCI is an alternative way of interacting with the environment by using brain activity instead of muscles and nerves. For that reason, BCI systems are of high clinical value for targeted populations suffering from neurological disorders. In this paper, we present a new processing approach in three publicly available BCI data sets: (a) a well-known multi-class (N = 6) coded-modulated Visual Evoked potential (c-VEP)-based BCI system for able-bodied and disabled subjects; (b) a multi-class (N = 32) c-VEP with slow and fast stimulus representation; and (c) a steady-state Visual Evoked potential (SSVEP) multi-class (N = 5) flickering BCI system. Estimating cross-frequency coupling (CFC) and namely δ-θ [δ: (0.5–4 Hz), θ: (4–8 Hz)] phase-to-amplitude coupling (PAC) within sensor and across experimental time, we succeeded in achieving high classification accuracy and Information Transfer Rates (ITR) in the three data sets. Our approach outperformed the originally presented ITR on the three data sets. The bit rates obtained for both the disabled and able-bodied subjects reached the fastest reported level of 324 bits/min with the PAC estimator. Additionally, our approach outperformed alternative signal features such as the relative power (29.73 bits/min) and raw time series analysis (24.93 bits/min) and also the original reported bit rates of 10–25 bits/min. In the second data set, we succeeded in achieving an average ITR of 124.40 ± 11.68 for the slow 60 Hz and an average ITR of 233.99 ± 15.75 for the fast 120 Hz. In the third data set, we succeeded in achieving an average ITR of 106.44 ± 8.94. Current methodology outperforms any previous methodologies applied to each of the three free available BCI datasets.

Shearlet and contourlet transforms for analysis of electrocardiogram signals

Background and Objective: Cardiac arrhythmia is an abnormal variation in the heart electrical activity that affects millions of people worldwide. Electrocardiogram (ECG) signals have been widely used to assess and diagnose cardiac abnormalities. Methods: A novel methodology based on shearlet and contourlet transforms for automatically classify an input ECG signal into different heart beat types is proposed and evaluated in this work. Classifiers are trained through a set of features extracted from these time-frequency coefficients. Results: Tests are conducted on MIT-BIH data set to demonstrate the effectiveness of the proposed classification method. The shearlet and contourlet transforms achieved high classification accuracy rates. Conclusions: The developed system can help cardiologists obtain structural and functional information of the heart by means of ECG patterns, improving their diagnostic tasks.

RoughPSO: rough set-based particle swarm optimisation

Particle swarm optimisation (PSO) is an optimisation algorithm based on stochastic search technique. PSO has many similar characteristics with evolutionary computation such as genetic algorithms (GA). Unlike GA, PSO has no evolution operators. In PSO, the particles (potential solutions) fly through the solution space by following the current optimum particles. However, PSO is easy to converge to a local optimum because the search process is stochastic. Rough set, in computer science, is a formal approximation of a conventional set in terms of a pair of sets. Rough set gives the lower and the upper approximation of the original set and is always used to deal with those uncertainty problems. In this paper, the properties of rough set theory are used to improve the local convergence problems in PSO, thereby an algorithm RoughPSO is proposed. RoughPSO utilises the lower- and upper-approximation sets of rough set to obtain the membership values. These values are then used to update the velocity and position of each particle. RoughPSO is applied for function optimisation and classification in machine learning. Empirical study shows that RoughPSO not only can solve the convergence to a local optimum, but also obtains higher classification accuracy rates on some datasets than those PSO-based classification algorithms.

Nasalization Amplitude-Timing Characteristics of Speakers With and Without Cleft Palate.

To examine the amplitude-temporal relationships of acoustic nasalization in speakers with a range of nasality and to determine the extent to which each domain independently predicts the speaker's perceived oral-nasal balance. Rate-controlled speech samples, consisting of /izinizi/, /azanaza/, and /uzunuzu/, were recorded from 18 participants (14 with repaired cleft palate and 4 without cleft palate) using the Nasometer. The mean nasalance of the entire mid-vowel-nasal consonant-vowel (mid-VNV) sequence (amplitude-domain) and the duration of the nasalized segment of the mid-VNV sequence (temporal-domain) were obtained based on nasalance contours. Strong linear and vowel-dependent relationships were observed between the 2 domains of nasalization (adjusted R2 = 71.5%). Both the amplitude- and temporal-domain measures were found to reliably predict the speaker's perceived oral-nasal balance, with better overall model fit and higher classification accuracy rates observed in /izinizi/ and /uzunuzu/ than in /azanaza/. Despite poor specificity, the temporal-domain measure of /azanaza/ was found to have a strong correlation with the participants' Zoo passage nasalance scores (rs = .897, p < .01), suggesting its potential utility as a severity indicator of perceived nasality. With the use of relatively simple speech tasks and measurements representing the amplitude and temporal domains of nasalization, the present study provided practical guidelines for using the Nasometer in assessing patients with oral-nasal resonance imbalance. Findings suggest that both domain measures of nasalization should be examined across different vowel contexts, given that each domain may provide clinically relevant, yet different, information.

A PROPOSAL OF KC-SVM CLASSIFIER FOR EEG FEATURES OBTAINED THROUGH SF FILTERING WITH APPLICATION ON BCI DATA

The objective of this paper is to classify Brain Computer Interfacing (BCI) competition III, dataset IIIA with low computational complexity and high classification accuracy rate. This study presents the investigation of self-filtering (SF) as a pre-processing technique and Wigner trispectrum as feature extraction technique followed by the proposed [Formula: see text]-means clustering SVM (KC-SVM) classifier. KC-SVM is an improved form of classifier, which comprises KC-SVM with 10-fold cross validation technique. To analyze each stage of classification, different evaluation indexes of respective methods are considered. Substantial improvement on the performance of proposed classification method is scrutinized. Performance measurement accuracy 98.1%, Kappa coefficient 0.94, mutual information of feature vectors 0.14 and Fisher score 6.5 are achieved for considered dataset. In summary, this paper demonstrates that the classification accuracy of proposed classifier together with SF and Wigner–Ville Trispectrum (WT) satisfies the required criteria of classification.

Digital forensics of microscopic images for printed source identification

When trying to identify a printed forged document, examining digital evidence can prove to be a challenge. In this study, microscopic images are used for printed source identification due to their high magnification properties resulting in detailed texture and structure information. Prior research implemented a scanner as a digitizing technique to resolve very fine printed identification, but this technique provided limited information on the resolution and magnification of the sample. In contrast, the performance of microscopy techniques can retrieve the shape and surface texture of a printed document with differing micro structures among printer sources. To explore the relationship between source printers and images obtained by the microscope, the proposed approach utilizes image processing techniques and data exploration methods to calculate many important statistical features, including: Local Binary Pattern (LBP), Gray Level Co-occurrence Matrix (GLCM), Discrete Wavelet Transform (DWT), Spatial filters, the Wiener filter, the Gabor filter, Haralick, and SFTA features. Among the different set of features, the LBP approach achieves the highest identification rate and is significantly superior to other methods. As a result, the proposed technique using microscopic images achieves a high classification accuracy rate, which shows promising applications for real world digital forensics research.

A multi-objective evolutionary algorithm-based soft computing model for educational data mining

Purpose The purpose of this paper is to propose a soft computing model based on multi-objective evolutionary algorithm (MOEA), namely, modified micro genetic algorithm (MmGA) coupled with a decision tree (DT)-based classifier, in classifying and optimising the students’ online interaction activities as classifier of student achievement. Subsequently, the results are transformed into useful information that may help educator in designing better learning instructions geared towards higher student achievement. Design/methodology/approach A soft computing model based on MOEA is proposed. It is tested on benchmark data pertaining to student activities and achievement obtained from the University of California at Irvine machine learning repository. Additional, a real-world case study in a distance learning institution, namely, Wawasan Open University in Malaysia has been conducted. The case study involves a total of 46 courses collected over 24 consecutive weeks with students across the entire regions in Malaysia and worldwide. Findings The proposed model obtains high classification accuracy rates at reduced number of features used. These results are transformed into useful information for the educational institution in our case study in an effort to improve student achievement. Whether benchmark or real-world case study, the proposed model successfully reduced the number features used by at least 48 per cent while achieving higher classification accuracy. Originality/value A soft computing model based on MOEA, namely, MmGA coupled with a DT-based classifier, in handling educational data is proposed.

Research on the Design of Parallel Hypernetworks Model Based on Hadoop Cloud Computing Platform

The evolutionary hypernetwork is a kind of cognitive learning model inspired by biological networks and the hypernetwork model has been successfully applied in plenty of fields. But the hypernetwork model may lead to a huge amount of hyperedges in the process of high dimensional data, which is the bottleneck of the single machine hardware implementation. In order to solve this problem, this paper presents the parallel model of hypernetwork computation under the Hadoop cloud computing platform, which changes single machine algorithm to multiple machines. Empirical studies on the acute leukemia dataset result demonstrate that the proposed hypernetwork classifier of parallel evolutionary hypernetwork model based on the Hadoop cloud computing platform has higher classification accuracy rate and good scalability. This is the main advantage of the parallel evolutionary hypernetwork model.

A double-elimination-tournament-based competitive co-evolutionary artificial neural network classifier

This paper presents a competitive co-evolutionary (ComCoE) that engages a double elimination tournament (DET) to evolve artificial neural networks (ANNs) for undertaking data classification problems. The proposed model performs a global search by a ComCoE approach to find near optimal solutions. During the global search process, two populations of different ANNs compete and fitness evaluation of each ANN is made in a subjective manner based on their participations throughout a DET which promotes competitive interactions among individual ANNs. The adaptation and fitness evaluation processes drive the global search for a more competent ANN classifier. A winning ANN is identified from the global search. Then, the Scaled Conjugate Backpropagation algorithm, which is a local search, is performed to further train the winning ANN to obtain a precise solution. The performance of the proposed classification model is evaluated rigorously; its performance is compared with the baseline ANNs of the proposed model as well as other classifiers. The results indicate that the proposed model could construct an ANN which could produce high classification accuracy rates with a compact network structure.

Identification of Anisomerous Motor Imagery EEG Signals Based on Complex Algorithms.

Motor imagery (MI) electroencephalograph (EEG) signals are widely applied in brain-computer interface (BCI). However, classified MI states are limited, and their classification accuracy rates are low because of the characteristics of nonlinearity and nonstationarity. This study proposes a novel MI pattern recognition system that is based on complex algorithms for classifying MI EEG signals. In electrooculogram (EOG) artifact preprocessing, band-pass filtering is performed to obtain the frequency band of MI-related signals, and then, canonical correlation analysis (CCA) combined with wavelet threshold denoising (WTD) is used for EOG artifact preprocessing. We propose a regularized common spatial pattern (R-CSP) algorithm for EEG feature extraction by incorporating the principle of generic learning. A new classifier combining the K-nearest neighbor (KNN) and support vector machine (SVM) approaches is used to classify four anisomerous states, namely, imaginary movements with the left hand, right foot, and right shoulder and the resting state. The highest classification accuracy rate is 92.5%, and the average classification accuracy rate is 87%. The proposed complex algorithm identification method can significantly improve the identification rate of the minority samples and the overall classification performance.