Optimal Classification Accuracy Research Articles

Feasibility of a Real-Time Clinical Augmented Reality and Artificial Intelligence Framework for Pain Detection and Localization From the Brain.

BackgroundFor many years, clinicians have been seeking for objective pain assessment solutions via neuroimaging techniques, focusing on the brain to detect human pain. Unfortunately, most of those techniques are not applicable in the clinical environment or lack accuracy.ObjectiveThis study aimed to test the feasibility of a mobile neuroimaging-based clinical augmented reality (AR) and artificial intelligence (AI) framework, CLARAi, for objective pain detection and also localization direct from the patient’s brain in real time.MethodsClinical dental pain was triggered in 21 patients by hypersensitive tooth stimulation with 20 consecutive descending cold stimulations (32°C-0°C). We used a portable optical neuroimaging technology, functional near-infrared spectroscopy, to gauge their cortical activity during evoked acute clinical pain. The data were decoded using a neural network (NN)–based AI algorithm to classify hemodynamic response data into pain and no-pain brain states in real time. We tested the performance of several networks (NN with 7 layers, 6 layers, 5 layers, 3 layers, recurrent NN, and long short-term memory network) upon reorganized data features on pain diction and localization in a simulated real-time environment. In addition, we also tested the feasibility of transmitting the neuroimaging data to an AR device, HoloLens, in the same simulated environment, allowing visualization of the ongoing cortical activity on a 3-dimensional brain template virtually plotted on the patients’ head during clinical consult.ResultsThe artificial neutral network (3-layer NN) achieved an optimal classification accuracy at 80.37% (126,000/156,680) for pain and no pain discrimination, with positive likelihood ratio (PLR) at 2.35. We further explored a 3-class localization task of left/right side pain and no-pain states, and convolutional NN-6 (6-layer NN) achieved highest classification accuracy at 74.23% (1040/1401) with PLR at 2.02.ConclusionsAdditional studies are needed to optimize and validate our prototype CLARAi framework for other pains and neurologic disorders. However, we presented an innovative and feasible neuroimaging-based AR/AI concept that can potentially transform the human brain into an objective target to visualize and precisely measure and localize pain in real time where it is most needed: in the doctor’s office.International Registered Report Identifier (IRRID)RR1-10.2196/13594

A novel measure of attribute significance with complexity weight

Attribute reduction is one of the most important problems in rough set theory. Conventional attribute reduction algorithms are based on minimal errors in seen objects, namely empirical risk minimization. Classification ability in unseen objects, namely generalization ability is more important in actual applications. Therefore, a good reduct should have good generalization ability. Structural risk minimization (SRM) inductive principle is an effective tool to control the generalization ability of learning machines, which considers complexity and errors in seen objects simultaneously. Therefore, this paper introduces the SRM principle into the definition of attribute significance, proposes that the number of rules can characterize the actual complexity of the rough set-based classifier effectively and defines a novel measure of attribute significance with complexity weight. Based on the new attribute significance, a new heuristic attribute reduction algorithm called HSRM-R algorithm is developed. The 10-fold cross-validation experiments in 21 UCI datasets show that HSRM-R algorithm obtains better generalization ability than conventional attribute reduction algorithms based on dependency degree, information entropy, Fisher score and Laplacian score. Further experiments show that HSRM-R algorithm obtains fewer rules and larger support coefficient. This means HSRM-R algorithm can extract stronger rules, which explains why it has better generalization ability to some extent. Although HSRM-R algorithm consumes more time than conventional algorithms, it obtains optimal classification accuracy in almost all datasets used in the experiments. Thus, the proposed HSRM-R algorithm provides an approach to guaranteeing the generalization ability theoretically in the case where users require high classification accuracy.

Structural risk minimization of rough set-based classifier

The classification ability in unseen objects, namely generalization ability, remains a long-standing challenge in rough set-based classifier. Current research mainly focuses on introducing thresholds to tolerate some errors in seen objects. The reason for introducing thresholds and the selection of threshold still lack sufficient theoretical support. The structural risk minimization (SRM) inductive principle is one of the most effective theories to control the generalization ability, which suggests a trade-off between errors in seen objects and complexity. Therefore, this paper introduces the SRM principle into rough set-based classifier and proposes SRM algorithm of rough set-based classifier called SRM-R algorithm. SRM-R algorithm uses the number of rules to characterize the actual complexity of rough set-based classifier and obtains the optimal trade-off between errors in seen objects and complexity through genetic multi-objective optimization. The tenfold cross-validation experiment in 12 UCI datasets shows SRM-R algorithm can significantly improve the generalization ability compared with conventional threshold algorithm. Besides, this paper uses other two possible complexity metrics including the number of attributes and attribute space to construct corresponding SRM algorithms, respectively, and compared their classification accuracy with that of SRM-R algorithm. Comparison result shows SRM-R algorithm obtains optimal classification accuracy. This indicates that the number of rules characterizes the complexity more effectively than the number of attributes and attribute space. Further experiments show that SRM-R algorithm obtains fewer rules and larger support coefficient, which means it extracts stronger rules. This explains why it obtains better generalization ability to some extent.

Feature Selection Techniques and Classification Accuracy of Supervised Machine Learning in Text Mining

Text mining is a special case of data mining which explore unstructured or semi-structured text documents, to establish valuable patterns and rules that indicate trends and significant features about specific topics. Text mining has been in pattern recognition, predictive studies, sentiment analysis and statistical theories in many areas of research, medicine, financial analysis, social life analysis, and business intelligence. Text mining uses concept of natural language processing and machine learning. Machine learning algorithms have been used and reported to give great results, but their performance of machine learning algorithms is affected by factors such as dataset domain, number of classes, length of the corpus, and feature selection techniques used. Redundant attribute affects the performance of the classification algorithm, but this can be reduced by using different feature selection techniques and dimensionality reduction techniques. Feature selection is a data preprocessing step that chooses a subset of input variable while eliminating features with little or no predictive information. Feature selection techniques are Information gain, Term Frequency, Term Frequency-Inverse document frequency, Mutual Information, and Chi-Square, which can use a filters, wrappers, or embedded approaches. To get the most value from machine learning, pairing the best algorithms with the right tools and processes is necessary. Little research has been done on the effect of feature selection techniques on classification accuracy for pairing of these algorithms with the best feature selection techniques for optimal results. In this research, a text classification experiment was conducted using incident management dataset, where incidents were classified into their resolver groups. Support vector machine (SVM), K-Nearest Neighbors (KNN), Naïve Bayes (NB) and Decision tree (DT) machine learning algorithms were examined. Filtering approach was used on the feature selection techniques, with different ranking indices applied for optimal feature set and classification accuracy results analyzed. The classification accuracy results obtained using TF were, 88% for SVM, 70% for NB, 79% for Decision tree, and KNN had 55%, while Boolean registered 90%, 83%, 82% and 75%, for SVM, NB, DT, and KNN respectively. TF-IDF, had 91%, 83%, 76%, and 56% for SVM, NB, DT, and KNN respectively. The results showed that algorithm performance is affected by feature selection technique applied. SVM performed best, followed by DT, KNN and finally NB. In conclusion, presence of noisy data leads to poor learning performance and increases the computational time. The classifiers performed differently depending on the feature selection technique applied. For optimal results, the classifier that performed best together with the feature selection technique with the best feature subset should be applied for all types of data for accurate classification performance. Keywords: Text Classification, Supervised Machine Learning, Feature Selection DOI : 10.7176/JIEA/9-3-06 Publication date :May 31 st 2019

A Novel Hybrid Algorithm for Feature Selection Based on Whale Optimization Algorithm

Feature selection enhances classification accuracy by removing irrelevant and redundant feature. Feature selection plays an important role in data mining and pattern recognition. In this paper, we propose a hybrid feature subset selection algorithm called the maximum Pearson maximum distance improved whale optimization algorithm (MPMDIWOA). First, based on Pearson's correlation coefficient and correlation distance, a filter algorithm is proposed named maximum Pearson maximum distance (MPMD). Two parameters are proposed in MPMD to adjust the weights of the relevance and redundancy. Second, the modified whale optimization algorithm can act as a wrapper algorithm. After introducing the maximum value without change (MVWC) and the threshold, the filter algorithm and the wrapper algorithm are combined to form an algorithm called MPMDIWOA. In MPMDIWOA, the filter algorithm and wrapper algorithm are called different times according to changes in MVWC and threshold. Finally, the optimal classification accuracy was found. The proposed method is tested on 10 benchmark datasets from UCI machine learning databases. The experimental results show that the classification accuracy of the proposed algorithm is significantly higher than that of the other three wrapper algorithms and one hybrid algorithm.

Notice of Violation of IEEE Publication Principles: Research on Urban Land Price Assessment Based on Artificial Neural Network Model

Notice of Violation of IEEE Publication Principles <br><br> “Research on Urban Land Price Assessment Based on Artificial Neural Network Model” <br> by Shousong Cai, Xiaomin Gu, Xiaoguang Wang, Ying Chen <br> in IEEE Access, Special Section on Urban Computing and Intelligence, December 2019, pp. 180738-180748 <br><br> After careful and considered review of the content and authorship of this paper by a duly constituted expert committee, this paper has been found to be in violation of IEEE’s Publication Principles. <br><br> This paper copied content from the paper cited below. The original content was translated and copied without attribution (including appropriate references to the original author(s) and/or paper title) and without permission. <br><br> “Research on Urban Residential Land Price Assessment Based on Artificial Intelligence Reasoning Model” <br> by Biao Zhang <br> in the Masters Thesis, University of Electronic Science and Technology of China, June 2018 <br><br> <br/> At present, there are many problems to be solved in the study of residential land price evaluation, such as the imbalance of residential land price data categories, the small sample of residential land price data sets, the possibility of the evaluation model falling into the local optimal value, and the large error of feature quantification. This paper introduces the problems in the process of residential land price evaluation from three aspects: land price data, evaluation model and feature quantification. In this paper, an extraction method for land price features based on transfer learning is proposed in connection with defects such as the small total amount of residential land price data and unbalanced class distribution, and different land price assessment models are used to carry out a precision evaluation and explore the models and methods suitable for land price assessment. In this paper, the extraction method for residential land price features is improved. The experimental results show that the extraction algorithm for land price features based on transfer learning has better classification accuracy than the commonly used principal component analysis and extraction method and linear normalization method, with a mean average precision 10.9% and 4.73% higher, respectively. The optimal feature set classification accuracy reached 90.28%; therefore, the algorithm for residential land price assessment proposed in this paper is able to satisfy the precision requirement of actual land price assessment.

Diagnosis and prognosis of Alzheimer's disease using brain morphometry and white matter connectomes.

Accurate, reliable prediction of risk for Alzheimer's disease (AD) is essential for early, disease-modifying therapeutics. Multimodal MRI, such as structural and diffusion MRI, is likely to contain complementary information of neurodegenerative processes in AD. Here we tested the utility of the multimodal MRI (T1-weighted structure and diffusion MRI), combined with high-throughput brain phenotyping—morphometry and structural connectomics—and machine learning, as a diagnostic tool for AD. We used, firstly, a clinical cohort at a dementia clinic (National Health Insurance Service-Ilsan Hospital [NHIS-IH]; N = 211; 110 AD, 64 mild cognitive impairment [MCI], and 37 cognitively normal with subjective memory complaints [SMC]) to test the diagnostic models; and, secondly, Alzheimer's Disease Neuroimaging Initiative (ADNI)-2 to test the generalizability. Our machine learning models trained on the morphometric and connectome estimates (number of features = 34,646) showed optimal classification accuracy (AD/SMC: 97% accuracy, MCI/SMC: 83% accuracy; AD/MCI: 97% accuracy) in NHIS-IH cohort, outperforming a benchmark model (FLAIR-based white matter hyperintensity volumes). In ADNI-2 data, the combined connectome and morphometry model showed similar or superior accuracies (AD/HC: 96%; MCI/HC: 70%; AD/MCI: 75% accuracy) compared with the CSF biomarker model (t-tau, p-tau, and Amyloid β, and ratios). In predicting MCI to AD progression in a smaller cohort of ADNI-2 (n = 60), the morphometry model showed similar performance with 69% accuracy compared with CSF biomarker model with 70% accuracy. Our comparisons of the classifiers trained on structural MRI, diffusion MRI, FLAIR, and CSF biomarkers showed the promising utility of the white matter structural connectomes in classifying AD and MCI in addition to the widely used structural MRI-based morphometry, when combined with machine learning.

An Artificial Intelligence (AI) Approach for Automated Flow Cytometric Diagnosis of B-Cell Lymphoma

Background: Multi-parameter flow cytometry (MFC) data is a diagnostic keystone for hematologic malignancies. Data analysis and interpretation requires resources and expertise, full automation is desired. Combination of automated gating strategies and machine learning on MFC data proved effective in detecting AML cell populations vs normal (Aghaeepour et al., 2013). Aim of our approach was to create tools capable of comprehensively analyzing and classifying MFC data. We developed a proof of principle framework to process MFC data generated from samples with suspected B-cell lymphomas.Methods: MFC data was obtained from routine diagnostics of 16,384 patients as follows: normal controls (n=8,493), chronic lymphocytic leukemia (CLL, n=3,412), CLL with increased prolymphocytes (CLL/PL, n=603), follicular lymphoma (FL, n=219), hairy cell leukemia (HCL, n=193), lymphoplasmacytic lymphoma (LPL, n=629), mantle cell lymphoma (MCL, n=269), marginal zone lymphoma (MZL, n=979), monoclonal B-cell lymphocytosis (MBL, n=1,480). Two 9-color combinations of monoclonal antibodies were applied to analyze surface expression of the following antigens besides scatter signals: FMC7, CD10, IgM, CD79b, CD20, CD23, CD19, CD5, CD45, Kappa, Lambda, CD38, CD25, CD11c, CD103, CD22. For dimensionality reduction we used an unsupervised clustering approach (Kohonen et al., 1990). We first generated a consensus self-organizing map (CS) of 100 nodes including all MFC events of multiple preselected cases (5 samples/cohort). For a test case the fluorescent profile of a cell was assigned to its nearest node in the CS (upsampling). We used the resulting distribution for classification by a sequential neural network (NN). Computationally intensive processing steps were run using Amazon Web Services. Mean accuracies were calculated counting individual cases across the entire dataset. All accuracies are described for 10 CS generations and 10-fold cross validation in the classification process. Approaches of data processing were sequentially developed aiming at maximization of classification accuracy.Results: First, we assessed the performance in a binary classification, i.e. the merged group of CLL and MBL (n=4,877) vs. normal (n=8,493) achieving 97% accuracy. Secondly, we applied our approach to the classification of all 9 classes with an accuracy of 74%. Visual inspection of tSNE (van der Maaten et al., 2008) transformed upsampling information revealed good separability between CLL and normal cases but substantial overlap between other lymphoma subtypes (Figure). The following approaches were applied to improve classification:1) Automated pregating of lymphocytes (CD45/SSC) by DBSCAN analysis (Ester et al., 1996) improved classification. Applying this to both CS generation and upsampling the accuracy of the 9-class problem increased to 78% (p<0.001).2) Using a CS trained only on cases with unequivocal phenotypes without pregating was significantly better (p=0.01), but pregating overrode this effect (p=0.5).3) According to the partial overlap of immunophenotypes hierarchical clustering of misclassifications showed overlap mainly between MCL and CLL/PL as well as between LPL and MZL. We therefore changed our approach merging the groups CLL/MBL, MCL/CLL/PL and LPL/MZL while keeping FL, HCL and normal controls separate. Automated classification of these 6 classes resulted in 86% accuracy.4) Further regrouping into 3 classes, according to the commonly applied routine approach of distinguishing normal cases from CD5+ (CLL, MBL, MCL, CLL/PL) and CD5- B-cell lymphoma (FL, LPL, MZL, HCL) resulted in a classification accuracy of 89%.5) The misclassified cases in the 3-class experiment had an average infiltration of 10% versus 41% for correctly classified ones. Limiting the automated classification to cases with ≥5% infiltration increased the classification accuracy to 95%.Conclusion: We created a framework for automated classification and analysis of routine diagnostic MFC data. Definition of entities and preselection of cell populations are essential for optimizing classification accuracy. Our approach is the first application of computational MFC and AI on a large scale dataset and proves the feasibility in a routine diagnostic work flow. To further improve the accuracy of classification a future algorithm will integrate distance metrics following SOM generation into the NN classifier. [Display omitted] DisclosuresHöllein:MLL Munich Leukemia Laboratory: Employment. Schabath:MLL Munich Leukemia Laboratory: Employment. Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Kern:MLL Munich Leukemia Laboratory: Employment, Equity Ownership.

Automated detection of vulnerable plaque in intravascular ultrasound images.

Acute coronary syndrome (ACS) is a syndrome caused by a decrease in blood flow in the coronary arteries. The ACS is usually related to coronary thrombosis and is primarily caused by plaque rupture followed by plaque erosion and calcified nodule. Thin-cap fibroatheroma (TCFA) is known to be the most similar lesion morphologically to a plaque rupture. In this paper, we propose methods to classify TCFA using various machine learning classifiers including feed-forward neural network (FNN), K-nearest neighbor (KNN), random forest (RF), and convolutional neural network (CNN) to figure out a classifier that shows optimal TCFA classification accuracy. In addition, we suggest pixel range-based feature extraction method to extract the ratio of pixels in the different region of interests to reflect the physician's TCFA discrimination criteria. Our feature extraction method examines the pixel distribution of the intravascular ultrasound (IVUS) image at a given ROI, which allows us to extract general characteristics of the IVUS image while simultaneously reflecting the different properties of the vessel's substances such as necrotic core and calcified nodule depending on the brightness of the pixel. A total of 12,325 IVUS images were labeled with corresponding optical coherence tomography (OCT) images to train and evaluate the classifiers. We achieved 0.859, 0.848, 0.844, and 0.911 area under the ROC curve (AUC) in the order of using FNN, KNN, RF, and CNN classifiers. As a result, the CNN classifier performed best and the top 10 features of the feature-based classifiers (FNN, KNN, RF) were found to be similar to the physician's TCFA diagnostic criteria. Graphical Abstract AUC result of proposed classifiers.

A hybrid feature selection algorithm for microarray data

For each microarray data set, only a small number of genes are beneficial. Due to the high-dimensional problem, gene selection research work remains a challenge. In order to solve the high-dimensional problem, we propose a dimensionality reduction algorithm named K value maximum relevance minimum redundancy improved grey wolf optimizer (KMR2IGWO). First, in the processing of KMR2, the K genes are selected. Second, the K genes are initialized by two ways according to random selection feature and different proportions of selection feature. Finally, the IGWO algorithm selects the optimal classification accuracy and the optimal combination of gene by adjusting the parameters of fitness function. The algorithm has a significant dimensionality reduction effect and is suitable for high-dimensional data sets. Experimental results show that the proposing KMR2IGWO strategy significantly reduces the dimension of microarray data and removes the redundant features. On the 14 microarray data sets, compared with the four algorithms mRMR + PSO, mRMR + GA, mRMR + BA, mRMR + CS, the proposed algorithm has higher performance in classification accuracy and feature subset length. In five data sets, the proposed algorithm average classification accuracy is 100%. On the 14 data sets, the proposed algorithm has a very significant dimensionality reduction effect, and the dimensionality reduction range is between 0.4% and 0.04%.

P 011—Towards an optimal combination of input signals and derived representations for gait classification based on ground reaction force measurements

Quantitative gait analysis produces a vast amount of data, which can be difficult to analyze. Automated gait classification based on machine learning techniques bear the potential to support clinicians in comprehending these complex data. Even though these techniques are already frequently used in the scientific community, there is no clear consensus on how the data need to be preprocessed and arranged to assure optimal classification accuracy outcomes.Is there an optimal data aggregation and preprocessing workflow to optimize classification accuracy outcomes?Based on our previous work on automated classification of ground reaction force (GRF) data, a sequential setup was followed: firstly, several aggregation methods – early fusion and late fusion – were compared, and secondly, based on the best aggregation method identified, the expressiveness of different combinations of signal representations was investigated. The employed dataset included data from 910 subjects, with four gait disorder classes and one healthy control group. The machine learning pipeline comprised principle component analysis (PCA), z-standardization and a support vector machine (SVM).The late fusion aggregation, i.e., utilizing majority voting on the classifier's predictions, performed best. In addition, the use of derived signal representations (relative changes and signal differences) seems to be advantageous as well.Our results indicate that great caution is needed when data preprocessing and aggregation methods are selected, as these can have an impact on classification accuracies. These results shall serve future studies as a guideline for the choice of data aggregation and preprocessing techniques to be employed.

Anatomical Biomarkers for Adolescent Major Depressive Disorder from Diffusion Weighted Imaging using SVM Classifier.

Adolescent Major Depressive Disorder (MDD) is a common and serious mental illness that could lead to tragic outcomes including chronic adult disability and suicide. In this paper, we explore anatomical features and apply machine learning approaches to identify responsive biomarkers distinguishing MDD patients from healthy subjects. The features of interest include metrics in two categories: a) anatomical connectivity defined by diffusion tensor imaging measurements between a pair of brain regions, and b) topological measurements from anatomical networks. A combination of p-value based filtering and minimum redundancy maximum relevance method is performed to select features for optimal classification accuracy. A leave-one-out cross-validation method is used for the classification performance evaluation. The proposed methodology achieves an improved accuracy of 78%, 90.39% sensitivity, and 79.66% precision for 79 subjects. The most distinguishing features are the betweenness centrality of the right lingual gyrus of the ADC network at 12% sparsity, the participation coefficient of the right lateral occipital sulcus of the ADC network at 22% sparsity, the participation coefficient of the right pars opercularis of the AD network at 16% sparsity, and the participation coefficient of the right lateral orbitofrontal cortex in the ADC network at 10% sparsity. Those network measures reflect the change of connectivity between the regions and their associated anatomical subnetworks.

An Intelligent Hybrid Feature Selection for Big Data Analytics:Application to Microarray Data Analysis

Recent advances in Genomics technologies has enabled cancer research to enter the Big Data era. In Big data analytics, both dimensionality reduction and designing an optimal analysis model is a challenging task. This paper presents an effective two phase feature selection model to effectively address this challenge in multi-class cancer classification problem. In the first phase, the model utilizes ensemble of filters to reduce the curse of dimensionality eliminating the irrelevant genes and maintaining the prognostic genes. In second phase, hybrid particle swarm optimization (HPSO) is employed to further reduce the dimensionality of prognostic genes and optimize the classifier structure synchronously to achieve optimal classification accuracy. To confirm the effectiveness of the proposed model, it is compared with other well-known recent methods on five benchmark microarray datasets using 10-fold cross validation. Experimental results demonstrates its superior performance in selecting compact prognostic gene subset while maintaining the classification accuracy.

Temporal Response Properties of Accessory Olfactory Bulb Neurons: Limitations and Opportunities for Decoding.

The vomeronasal system (VNS) is a major vertebrate chemosensory system that functions in parallel to the main olfactory system (MOS). Despite many similarities, the two systems dramatically differ in the temporal domain. While MOS responses are governed by breathing and follow a subsecond temporal scale, VNS responses are uncoupled from breathing and evolve over seconds. This suggests that the contribution of response dynamics to stimulus information will differ between these systems. While temporal dynamics in the MOS are widely investigated, similar analyses in the accessory olfactory bulb (AOB) are lacking. Here, we have addressed this issue using controlled stimulus delivery to the vomeronasal organ of male and female mice. We first analyzed the temporal properties of AOB projection neurons and demonstrated that neurons display prolonged, variable, and neuron-specific characteristics. We then analyzed various decoding schemes using AOB population responses. We showed that compared with the simplest scheme (i.e., integration of spike counts over the entire response period), the division of this period into smaller temporal bins actually yields poorer decoding accuracy. However, optimal classification accuracy can be achieved well before the end of the response period by integrating spike counts within temporally defined windows. Since VNS stimulus uptake is variable, we analyzed decoding using limited information about stimulus uptake time, and showed that with enough neurons, such time-invariant decoding is feasible. Finally, we conducted simulations that demonstrated that, unlike the main olfactory bulb, the temporal features of AOB neurons disfavor decoding with high temporal accuracy, and, rather, support decoding without precise knowledge of stimulus uptake time.SIGNIFICANCE STATEMENT A key goal in sensory system research is to identify which metrics of neuronal activity are relevant for decoding stimulus features. Here, we describe the first systematic analysis of temporal coding in the vomeronasal system (VNS), a chemosensory system devoted to socially relevant cues. Compared with the main olfactory system, timescales of VNS function are inherently slower and variable. Using various analyses of real and simulated data, we show that the consideration of response times relative to stimulus uptake can aid the decoding of stimulus information from neuronal activity. However, response properties of accessory olfactory bulb neurons favor decoding schemes that do not rely on the precise timing of stimulus uptake. Such schemes are consistent with the variable nature of VNS stimulus uptake.

Eye state classification from electroencephalography recordings using machine learning algorithms

Background and Objectives: Current developments in electroencephalography (EEG) foster medical and nonmedical applications outside the hospitals. For example, continuous monitoring of mental and cognitive states can contribute to avoid critical and potentially dangerous situations in daily life. An important prerequisite for successful EEG at home is a real-time classification of mental states. In this article, we compare different machine learning algorithms for the classification of eye states based on EEG recordings. Materials and Methods: We tested 23 machine learning algorithms from the Waikato Environment for Knowledge Analysis toolkit. Each classifier was analyzed on four different datasets, since two separate approaches - called sample-wise and segment-wise - in combination with raw and filtered data were applied. These datasets were recorded for 27 volunteers. The different approaches are compared in terms of accuracy, complexity, training time, and classification time. Results: Ten out of 23 classifiers fulfilled the determined requirements of high classification accuracy and short time of classification and can be denoted as applicable for real-time EEG eye state classification. Conclusions: We found that it is possible to predict eye states using EEG recordings with an accuracy from about 96% to over 99% in a real-time system. On the other hand, we found no best, universal method of classifying EEG eye states in all volunteers. Therefore, we conclude that the best algorithm should be chosen individually, using the optimal classification accuracy in combination with time of classification as the criterion.

Analysis of fluoroquinolones antibiotic residue in feed matrices using terahertz spectroscopy.

As antibiotic residue becomes more and more serious all over the world, a rapid and effective detection method is needed to evaluate the antibiotic residue in feed matrices to ensure food safety for consumers. In this study, three different kinds of fluoroquinolones (norfloxacin, enrofloxacin, and ofloxacin) in feed matrices were analyzed using terahertz (THz) spectroscopy, respectively. Meanwhile, pure fluoroquinolones and pure feed matrices were also measured in the same way. Then, the absorption spectra of all of the samples were extracted in the transmission mode. Pure norfloxacin has two absorption peaks at 0.825 and 1.187THz, and they could still be observed when mixing norfloxacin with feed matrices. Also, there was an obvious and strong absorption peak for ofloxacin at 1.044THz. However, no obvious absorption peak for enrofloxacin was observed, and only a weak absorption peak was located at 0.8THz. Then, the different models were established with different chemometrics to identify the fluoroquinolones in feed matrices and determined the fluoroquinolones content in the feed matrices. The least squares support vector machines, Naive Bayes, Mahalanobis distance, and back propagation neural network (BPNN) were used to build the identification model with a Savitzky-Golay filter and standardized normal variate pretreatments. The results show that the excellent classification model was acquired with the BPNN combined with no pretreatment. The optimal classification accuracy was 80.56% in the testing set. After that, multiple linear regression and stepwise regression were used to establish the quantitative detection model for different kinds of fluoroquinolones in feed matrices. The optimal correlation coefficients for norfloxacin, enrofloxacin, and ofloxacin in the prediction set were obtained with multiple linear regression that combined absorption peaks with wavelengths selected by stepwise regression, which were 0.867, 0.828, and 0.964, respectively. Overall, this research explored the potential of identifying the fluoroquinolones in feed matrices using THz spectroscopy without a complex pretreatment process and then quantitatively detecting the fluoroquinolones content in feed matrices. The results demonstrate that THz spectra could be used to identify fluoroquinolones in feed matrices and also detect their content quantitatively, which has great significance for the food safety industry.

A New Approach for Improving Classification Accuracy in Predictive Discriminant Analysis

The focus of a predictive discriminant analysis is to improve classification accuracy, and to obtain statistically optimal classification accuracy or hit rate is still a challenge due to the inherent variability of most real life dataset. Improving classification accuracy is usually achieved with best subset of relevant predictors obtained by using classical variable selection methods. The goal of variable selection methods is to choose the best subset (or training sample) of relevant variables that typically reduces the complexity of a model and makes it easier to interpret, improves the classification accuracy of the model and reduces the training time. However, a statistically optimal hit rate can be achieved if the training sample meets a near optimal condition by resolving any significant differences in the variances for the groups formed by the dependent variable. This paper proposes a new approach for obtaining a near optimal training sample that will produce a statistically optimal hit rate using a modified winsorization with graphical diagnostic. In application to real life data sets, the proposed new approach was able to identify and remove legitimate contaminants in one or more predictors in the training sample, thereby resolving any significant differences in the variances for the groups formed by the dependent variable. The graphical diagnostic associated with the new approach, however, provides a useful visual tool which served as an alternative graphical test for homogeneity of variances.

Acquiring and preprocessing leaf images for automated plant identification: understanding the tradeoff between effort and information gain

BackgroundAutomated species identification is a long term research subject. Contrary to flowers and fruits, leaves are available throughout most of the year. Offering margin and texture to characterize a species, they are the most studied organ for automated identification. Substantially matured machine learning techniques generate the need for more training data (aka leaf images). Researchers as well as enthusiasts miss guidance on how to acquire suitable training images in an efficient way.MethodsIn this paper, we systematically study nine image types and three preprocessing strategies. Image types vary in terms of in-situ image recording conditions: perspective, illumination, and background, while the preprocessing strategies compare non-preprocessed, cropped, and segmented images to each other. Per image type-preprocessing combination, we also quantify the manual effort required for their implementation. We extract image features using a convolutional neural network, classify species using the resulting feature vectors and discuss classification accuracy in relation to the required effort per combination.ResultsThe most effective, non-destructive way to record herbaceous leaves is to take an image of the leaf’s top side. We yield the highest classification accuracy using destructive back light images, i.e., holding the plucked leaf against the sky for image acquisition. Cropping the image to the leaf’s boundary substantially improves accuracy, while precise segmentation yields similar accuracy at a substantially higher effort. The permanent use or disuse of a flash light has negligible effects. Imaging the typically stronger textured backside of a leaf does not result in higher accuracy, but notably increases the acquisition cost.ConclusionsIn conclusion, the way in which leaf images are acquired and preprocessed does have a substantial effect on the accuracy of the classifier trained on them. For the first time, this study provides a systematic guideline allowing researchers to spend available acquisition resources wisely while yielding the optimal classification accuracy.

Dynamical recursive feature elimination technique for neurophysiological signal-based emotion recognition

The machine learning-based classification model can predict the operator emotional states in human–machine system based on nonlinear, multidimensional neurophysiological features. However, the dynamical properties of the testing physiological data regarding the time series may influence the feature distribution variation and inter-class discrimination across different time steps. To overcome this shortcoming, we propose a novel EEG feature selection method, dynamical recursive feature elimination (D-RFE), to find the optimal but different feature rankings at each time instant for arousal and valence recognition. With the classification framework implemented via a model-selected least square support vector machine, the participant-specific classification performance has been significantly improved against conventional RFE model and several common classifiers. The optimal classification accuracy and F1-score elicited by the proposed method are 0.7896, 0.7991, 0.7143, and 0.7257 for arousal and valence dimensions, respectively, which are quite competitive among recent reported works on the same EEG database.

The impact of training class proportions on binary cropland classification

ABSTRACTThe ground truth data sets required to train supervised classifiers are usually collected as to maximize the number of samples under time, budget and accessibility constraints. Yet, the performance of machine learning classifiers is, among other factors, sensitive to the class proportions of the training set. In this letter, the joint effect of the number of calibration samples and the class proportions on the accuracy was systematically quantified using two state-of-the-art machine learning classifiers (random forests and support vector machines). The analysis was applied in the context of binary cropland classification and focused on two contrasted agricultural landscapes. Results showed that the classifiers were more sensitive to class proportions than to sample size, though sample size had to reach 2,000 pixels before its effect leveled off. Optimal accuracies were obtained when the training class proportions were close to those actually observed on the ground. Then, synthetic minority over-sampling technique (SMOTE) was implemented to artificially regenerate the native class proportions in the training set. This resampling method led to an increase of the accuracy of up to 30%. These results have direct implications for (i) informing data collection strategies and (ii) optimizing classification accuracy. Though derived for cropland mapping, the recommendations are generic to the problem of binary classification.