Committee Of Classifiers Research Articles

Deep Learning for Combating Misinformation in Multicategorical Text Contents.

Currently, one can observe the evolution of social media networks. In particular, humans are faced with the fact that, often, the opinion of an expert is as important and significant as the opinion of a non-expert. It is possible to observe changes and processes in traditional media that reduce the role of a conventional 'editorial office', placing gradual emphasis on the remote work of journalists and forcing increasingly frequent use of online sources rather than actual reporting work. As a result, social media has become an element of state security, as disinformation and fake news produced by malicious actors can manipulate readers, creating unnecessary debate on topics organically irrelevant to society. This causes a cascading effect, fear of citizens, and eventually threats to the state's security. Advanced data sensors and deep machine learning methods have great potential to enable the creation of effective tools for combating the fake news problem. However, these solutions often need better model generalization in the real world due to data deficits. In this paper, we propose an innovative solution involving a committee of classifiers in order to tackle the fake news detection challenge. In that regard, we introduce a diverse set of base models, each independently trained on sub-corpora with unique characteristics. In particular, we use multi-label text category classification, which helps formulate an ensemble. The experiments were conducted on six different benchmark datasets. The results are promising and open the field for further research.

Quality-Aware Self-Training on Differentiable Synthesis of Rare Relational Data

Data scarcity is a very common real-world problem that poses a major challenge to data-driven analytics. Although a lot of data-balancing approaches have been proposed to mitigate this problem, they may drop some useful information or fall into the overfitting problem. Generative Adversarial Network (GAN) based data synthesis methods can alleviate such a problem but lack of quality control over the generated samples. Moreover, the latent associations between the attribute set and the class labels in a relational data cannot be easily captured by a vanilla GAN. In light of this, we introduce an end-to-end self-training scheme (namely, Quality-Aware Self-Training) for rare relational data synthesis, which generates labeled synthetic data via pseudo labeling on GAN-based synthesis. We design a semantic pseudo labeling module to first control the quality of the generated features/samples, then calibrate their semantic labels via a classifier committee consisting of multiple pre-trained shallow classifiers. The high-confident generated samples with calibrated pseudo labels are then fed into a semantic classification network as augmented samples for self-training. We conduct extensive experiments on 20 benchmark datasets of different domains, including 14 industrial datasets. The results show that our method significantly outperforms state-of-the-art methods, including two recent GAN-based data synthesis schemes. Codes are available at https://github.com/yaxinhou/QAST.

Contextually Enriched Meta-Learning Ensemble Model for Urdu Sentiment Analysis

The task of analyzing sentiment has been extensively researched for a variety of languages. However, due to a dearth of readily available Natural Language Processing methods, Urdu sentiment analysis still necessitates additional study by academics. When it comes to text processing, Urdu has a lot to offer because of its rich morphological structure. The most difficult aspect is determining the optimal classifier. Several studies have incorporated ensemble learning into their methodology to boost performance by decreasing error rates and preventing overfitting. However, the baseline classifiers and the fusion procedure limit the performance of the ensemble approaches. This research made several contributions to incorporate the symmetries concept into the deep learning model and architecture: firstly, it presents a new meta-learning ensemble method for fusing basic machine learning and deep learning models utilizing two tiers of meta-classifiers for Urdu. The proposed ensemble technique combines the predictions of both the inter- and intra-committee classifiers on two separate levels. Secondly, a comparison is made between the performance of various committees of deep baseline classifiers and the performance of the suggested ensemble Model. Finally, the study’s findings are expanded upon by contrasting the proposed ensemble approach efficiency with that of other, more advanced ensemble techniques. Additionally, the proposed model reduces complexity, and overfitting in the training process. The results show that the classification accuracy of the baseline deep models is greatly enhanced by the proposed MLE approach.

Computer-aided classification of successional stage in subtropical Atlantic Forest: a proposal based on fuzzy artificial intelligence

Due to the continuous variability of the forest regeneration process, patterns of indicator variables with membership in more than one successional stage may occur, making the classification of such stages a challenging and complex task. This study aims at presenting a comparative analysis of artificial intelligence methods as an alternative for computer-aided classification of successional stages in subtropical Atlantic Forest. As a research hypothesis, the authors consider that a fuzzy inference system should provide the best performance due to its ability to deal with uncertainties inherent to complex processes. The analyses were carried out using a database of the forest inventory of Santa Catarina, Southern Brazil. The data are composed of 177 sampling units of subtropical Atlantic Forest (mixed ombrophilous forest), characterized according to eighth indicator variables verified from the field by experts. This database was employed to train several machine learning methods under a tenfold cross-validation process. The overall accuracy (θ) and kappa coefficient were used to compare the performance between FIS and neural networks, classifier committees and support vector machine. Then, to verify if the classification by the FIS differed from the one performed by experts, the Kappa index and a statistical significance analysis by Pearson's [Formula: see text] test were determined. The hypotheses were verified with two-way tests at a significance level (α) 0.05, for a test power (1-β) 0.8 and minimum expected effect size between medium (ρ = 0.3). Statistical significance tests confirmed the hypothesis that FIS achieved the highest performance, with θ = 98.3% and a kappa value equal to 0.93 (almost perfect agreement) and showed no significant difference ([Formula: see text] = 0.047, p = 0.976) in comparison with the classification by experts. The use of FIS represents a promising alternative as a tool applicable for computer-aided classification of successional stages in subtropical Atlantic Forest. The results and conclusions should substantially impact the guidelines and decision-making process for deforestation authorizations and applicable compensation measures, which are based on the forest succession stage.

Identification of human mental workload levels in a language comprehension task with imbalance neurophysiological data

Background and ObjectiveOperator's capability for accurately comprehending verbal commands is critically important to maintain the performance of human-machine interaction. It can be evaluated by human mental workload measured with electroencephalography (EEG). However, the time duration of different workload conditions within a task session is unequal due to varied psychophysiological processes across individuals. It leads to data imbalance of the EEG for training workload classifiers. MethodsIn this study, we propose an EEG feature oversampling technique, Gaussian-SMOTE based feature ensemble (GSMOTE-FE), for workload recognition with imbalanced classes. First, artificial EEG instances are drawn from a Gaussian distribution in the margin between the minority and majority workload classes. Tomek links are detected as clues to remove redundant feature vectors. Then, we embed a feature selection module based on the GINI importance while an ensemble classifier committee with bootstrap aggregating is used to further enhance classification performance. ResultsWe validate the GSMOTE-FE framework based on an experiment that simulates operators to understand the correct meaning of the instructions in the Chinese language. Participants’ EEG signals and reaction time data were both recorded to validate the proposed workload classifier. Workload classification accuracy and Macro-F1 values are 0.6553 and 0.5862, respectively. Corresponding G-mean and AUC achieve at 0.5757 and 0.5958, respectively. ConclusionsThe performance of the GSMOTE-FE is demonstrated to be comparable with the advanced oversampling techniques. The workload classifier has the capability to indicate low and high levels of the task demand of the Chinese language understanding task.

Subspace-based decision trees integration

The ensemble methods are meta-algorithms that combine several base machine learning techniques to produce one optimal predictive model. Many existing committees of classifiers use feature space partitioning to determine the decision regions in which the selection of base classifiers occur or to learn a diverse ensemble of base classifiers. Furthermore, the division of the feature space into subspaces does not depend on the predictive model that defines the decision boundaries. Therefore, we propose a novel ensemble learning algorithm based on the feature space partitioning defined by decision tree boundaries. In addition, in a proposed framework, the feature subspace selection occurs taking class label imbalance ratio into account. The proposed method defines the ensemble class label based on selected previously feature subspaces and a neighborhood of feature subspaces defined by its reference point. Experimental results indicate that our proposed method is more effective than state-of-the-art ensemble methods on twenty-seven benchmark datasets regarding seven representative classification performance measures.

In-Field Detection of American Foulbrood (AFB) by Electric Nose Using Classical Classification Techniques and Sequential Neural Networks.

American foulbrood is a dangerous bee disease that attacks the sealed brood. It quickly leads to the death of bee colonies. Efficient diagnosis of this disease is essential. As specific odours are produced when larvae rot, it was investigated whether an electronic nose can distinguish between colonies affected by American foulbrood and healthy ones. The experiment was conducted in an apiary with 18 bee families, 9 of which showed symptoms of the disease confirmed by laboratory diagnostics. Three units of the Beesensor V.2 device based on an array of six semiconductor TGS gas sensors, manufactured by Figaro, were tested. Each copy of the device was tested in all bee colonies: sick and healthy. The measurement session per bee colony lasted 40 min and yielded results from four 10 min measurements. One 10-min measurement consisted of a 5 min regeneration phase and a 5 min object-measurement phase. For the experiments, we used both classical classification methods such as k-nearest neighbour, Naive Bayes, Support Vector Machine, discretized logistic regression, random forests, and committee of classifiers, that is, methods based on extracted representative data fragments. We also used methods based on the entire 600 s series, in this study of sequential neural networks. We considered, in this study, six options for data preparation as part of the transformation of data series into representative results. Among others, we used single stabilised sensor readings as well as average values from stable areas. For verifying the quality of the classical classifiers, we used the 25-fold train-and-test method. The effectiveness of the tested methods reached a threshold of 75 per cent, with results stable between 65 and 70 per cent. As an element to confirm the possibility of class separation using an artificial nose, we used applied visualisations of classes. It is clear from the experiments conducted that the artificial nose tested has practical potential. Our experiments show that the approach to the problem under study by sequential network learning on a sequence of data is comparable to the best classical methods based on discrete data samples. The results of the experiment showed that the Beesensor V.2 along with properly selected classification techniques can become a tool to facilitate rapid diagnosis of American foulbrood under field conditions.

An effective ensemble deep learning framework for text classification

Over the last decade Deep learning-based models surpasses classical machine learning models in a variety of text classification tasks. The primary challenge with text classification is determining the most appropriate deep learning classifier. Numerous research initiatives incorporated ensemble learning to boost the performance, minimize errors and avoid overfitting. However, the performance of the ensemble-methods is limited by the baseline classifiers and the fusion method. The current study makes the following contributions: First, it proposes a new meta-learning ensemble method that fuses baseline deep learning models using 2-tiers of meta-classifiers. Second, it conducts several experiments on six public benchmark datasets to evaluate the performance of the proposed ensemble. For each benchmark dataset, committees of different deep baseline classifiers are trained, and their best performance is compared with the performance of the proposed ensemble. Furthermore, the paper extends the results by comparing the performance of the proposed ensemble method to other state-of-the-art ensemble methods. The findings indicate that the proposed ensemble method significantly improve the classification accuracy of the baseline deep models. Furthermore, the proposed method outperforms the state-of-art ensemble methods. Finally, using the probability distributions for each class label of the deep baseline models improves the performance of the proposed ensemble method.

Network traffic analysis based on machine learning methods

Comparison of mining algorithms in the problem of detecting malicious network activity based on machine learning models is performed. A structural diagram of a system for analyzing network traffic in an industrial network based on machine learning methods has been developed. On one of the known datasets (CICIDS17), a series of experiments was carried out on preliminary analysis and preprocessing of features, highlighting the most significant features and building final models of classifiers. The f1-measure score for the committee of classifiers on the test sample is 0.967.

Customizable and committee data mining framework for stock trading

This study proposes a customizable and committee data mining (DM) framework for stock trading. The underlying conjecture is that different stocks may have different trading properties. A well-performing DM algorithm for a particular stock may not work well for other stocks. Therefore, the trading method designed for a particular stock should be customizable. To systematically obtain a customizable solution, we propose a research framework that can accommodate a large number of trading classifiers. Then, we select the well-performing ones to form a committee to make trading decisions for a particular stock. A committee decision is made because the best-performing trading classifier in a particular scenario (called validation phase) may not be the best one in another scenario (called testing phase). Therefore, some other well-performing classifiers afford their own advantages. We therefore extend the ensemble concept to form a customizable committee of well-performing classifiers that makes a trading decision by voting. To justify the idea of a customizable committee classifier, we slightly modify the framework and develop another trading method called customizable champion classifier in which only the best-performing classifier in the validation phase is used for stock trading. Using 12 stocks listed in the Taiwan Stock Exchange (TWSE) as test examples, the proposed customizable committee method outperforms the customizable champion method as well as six benchmarks methods (the buy-and-hold strategy and five popular technical analysis methods). To fairly compare a stock-trading method against the buy-and-hold strategy, two issues that have been rarely noted in literature are addressed in this study.

An abstract model of an artificial immune network based on a classifier committee for biometric pattern recognition by the example of keystroke dynamics

An abstract model of an artificial immune network (AIS) based on a classifier committee and robust learning algorithms (with and without a teacher) for classification problems, which are characterized by small volumes and low representativeness of training samples, are proposed. Evaluation of the effectiveness of the model and algorithms is carried out by the example of the authentication task using keyboard handwriting using 3 databases of biometric metrics. The AIS developed possesses emergence, memory, double plasticity, and stability of learning. Experiments have shown that AIS gives a smaller or comparable percentage of errors with a much smaller training sample than neural networks with certain architectures.

Locally robust EEG feature selection for individual-independent emotion recognition

Brain computer interface (BCI) systems can decode brain affective activities into interpretable features and facilitate emotional human–computer interaction. However, individual differences of neurophysiological responses from BCI subjects constitute a stumbling block in individual-independent emotion recognition. In this study, we propose a new locally-robust feature selection (LRFS) method to determine generalizable features of electroencephalography (EEG) within several subsets of accessible subjects. In the LRFS framework, extracted EEG features are first modeled with probability densities. By evaluating the similarity of all density functions between each two subjects, inter-individual consistency of the EEG features is described. The derived consistency determines locally-robust EEG features, wherein importance of each feature is examined according to margin loss between emotions. To fuse selected features from multiple subsets of subjects, we employ ensemble learning principle and build an emotion classifier committee. Based on public DEAP and MAHNOB-HCI databases, individual-independent classification accuracy of the LRFS-based classifier is achieved by 0.65–0.68 (DEAP) and 0.67–0.70 (MAHNOB-HCI) for arousal and valence domains, respectively. Competitiveness of the LRFS has been validated when compared with several existing feature selection methods and emotion recognition systems.

Emotional Speech Recognition Based on the Committee of Classifiers

This article presents the novel method for emotion recognition from speech based on committee of classifiers. Different classification methods were juxtaposed in order to compare several alternative approaches for final voting. The research is conducted on three different types of Polish emotional speech: acted out with the same content, acted out with different content, and spontaneous. A pool of descriptors, commonly utilized for emotional speech recognition, expanded with sets of various perceptual coefficients, is used as input features. This research shows that presented approach improve the performance with respect to a single classifier.

Learning concept drift with ensembles of optimum-path forest-based classifiers

Concept drift methods learn patterns in non-stationary environments. Although such behavior is usually not expected in traditional classification problems, in real-world scenarios one can face them very much easier. In such a context, classifiers can be fooled and their effectiveness affected as well. Some examples include theft detection in energy distribution systems, where the consumer’s behavior may change suddenly or smoothly, or even churn prediction in mobile companies. In this paper, we introduce the Optimum-Path Forest (OPF) classifier in the context of concept drift, using decisions for concept drift handling based on a committee of OPF classifiers. We consider three distinct perspectives (three rounds of experiments with variations of streaming managements) over publics datasets, being the results compared to the ones obtained by standard OPF. We consider OPF ensemble suitable to work under these dynamic scenarios since its recognition rates were considerably better when compared to traditional OPF.

An ensemble approach for supporting the respiratory isolation of presumed tuberculosis inpatients

Tuberculosis remains a global health challenge, especially in low and middle-income countries. New diagnostic tools can allow earlier diagnosis, reducing both the mortality and transmission in the community. In hospitals, the decision making relative to the allocation of presumed pulmonary tuberculosis inpatients in airborne rooms is critical, since no standard criterion has been established. In this paper, we propose a novel technique for developing a committee of classifiers aiming at supporting the decision making relative to inpatient respiratory isolation. The proposed approach is agnostic on the classification model adopted, exploiting tailored strategies for optimally integrating a small and diverse set of compact classifiers, resulting in highly accurate committees. The results confirm that the resulting committees have outperformed several recently proposed single-models and ensemble solutions, including deep learning techniques. As a practical benefit, the adoption of such decision support tool can reduce in almost one half the percentage of inpatients unnecessarily isolated at a university hospital.

Evidential query-by-committee active learning for pedestrian detection in high-density crowds

The automatic detection of pedestrians in dense crowds has become recently a very active topic of research due to the implications for public safety, and also due to the increased frequency of large scale social events. The detection task is complicated by multiple factors such as strong occlusions, high homogeneity, small target size, etc., and different types of detectors are able to provide complementary interpretations of the input data, with varying individual levels of performance. Our first contribution consists in outlining a fusion strategy under the form of an ensemble method, which models the imprecision arising from each of the detectors, both in the calibration and in the spatial domains in an evidential framework. Then, we propose a sample selection for augmenting the training set used jointly by the committee of classifiers, based on evidential disagreement measures among the base members in a Query-by-Committee context. The results show that the proposed fusion algorithm is effective in exploiting the strengths of the individual classifiers, as well as in augmenting the training set with informative samples which allow the resulting detector to enhance its performance.

Analysing user sentiment of Indian movie reviews

Purpose To be sustainable and competitive in the current business environment, it is useful to understand users’ sentiment towards products and services. This critical task can be achieved via natural language processing and machine learning classifiers. This paper aims to propose a novel probabilistic committee selection classifier (PCC) to analyse and classify the sentiment polarities of movie reviews. Design/methodology/approach An Indian movie review corpus is assembled for this study. Another publicly available movie review polarity corpus is also involved with regard to validating the results. The greedy stepwise search method is used to extract the features/words of the reviews. The performance of the proposed classifier is measured using different metrics, such as F-measure, false positive rate, receiver operating characteristic (ROC) curve and training time. Further, the proposed classifier is compared with other popular machine-learning classifiers, such as Bayesian, Naïve Bayes, Decision Tree (J48), Support Vector Machine and Random Forest. Findings The results of this study show that the proposed classifier is good at predicting the positive or negative polarity of movie reviews. Its performance accuracy and the value of the ROC curve of the PCC is found to be the most suitable of all other classifiers tested in this study. This classifier is also found to be efficient at identifying positive sentiments of reviews, where it gives low false positive rates for both the Indian Movie Review and Review Polarity corpora used in this study. The training time of the proposed classifier is found to be slightly higher than that of Bayesian, Naïve Bayes and J48. Research limitations/implications Only movie review sentiments written in English are considered. In addition, the proposed committee selection classifier is prepared only using the committee of probabilistic classifiers; however, other classifier committees can also be built, tested and compared with the present experiment scenario. Practical implications In this paper, a novel probabilistic approach is proposed and used for classifying movie reviews, and is found to be highly effective in comparison with other state-of-the-art classifiers. This classifier may be tested for different applications and may provide new insights for developers and researchers. Social implications The proposed PCC may be used to classify different product reviews, and hence may be beneficial to organizations to justify users’ reviews about specific products or services. By using authentic positive and negative sentiments of users, the credibility of the specific product, service or event may be enhanced. PCC may also be applied to other applications, such as spam detection, blog mining, news mining and various other data-mining applications. Originality/value The constructed PCC is novel and was tested on Indian movie review data.

Neural Networks in Video-Based Age and Gender Recognition on Mobile Platforms

The paper considers the use of convolutional neural networks for the concurrent recognition of the gender and age of a person by video records of his face. The emphasis is on the incorporation of the approach into mobile video analytics systems. We have investigated the fusion of decisions obtained during the processing of each video frame, including the use of the classifier committee based on Dempster-Shafer theory. We propose the novel age prediction method using the evaluation of the expectation of the most probable ages. We have compared existing neural-net models with a specially trained modification of the MobileNet convolution network with two outputs. The experimental results are given for such data collections as Kinect, IJB-A, Indian Movie and EmotiW. As compared with other conventional methods, our approach makes it possible to increase the age and gender recognition accuracy by 2–5% and 5–10% respectively.

An ensemble learning approach to lip-based biometric verification, with a dynamic selection of classifiers

Machine learning approaches are largely focused on pattern or object classification, where a combination of several classifier systems can be integrated to help generate an optimal or suboptimal classification decision. Multiple classification systems have been extensively developed because a committee of classifiers, also known as an ensemble, can outperform the ensemble’s individual members. In this paper, a classification method based on an ensemble of binary classifiers is proposed. Our strategy consists of two phases: (1) the competence of the base heterogeneous classifiers in a pool is determined, and (2) an ensemble is formed by combining those base classifiers with the greatest competences for the given input data.We have shown that the competence of the base classifiers can be successfully calculated even if the number of their learning examples was limited. Such a situation is particularly observed with biometric data. In this paper, we propose a new biometric data structure, the Sim coefficients, along with an efficient data processing technique involving a pool of competent classifiers chosen by dynamic selection.

Using a classifier ensemble for proactive quality monitoring and control: The impact of the choice of classifiers types, selection criterion, and fusion process

In recent times, the manufacturing processes are faced with many external or internal (the increase of customized product re-scheduling, process reliability…) changes. Therefore, monitoring and quality management activities for these manufacturing processes are difficult. Thus, the managers need more proactive approaches to deal with this variability. In this study, a proactive quality monitoring and control approach based on classifiers to predict defect occurrences and provide optimal values for factors critical to the quality processes is proposed. In a previous work (Noyel et al., 2013), the classification approach had been used in order to improve the quality of a lacquering process at a company plant; the results obtained are promising, but the accuracy of the classification model used needs to be improved. One way to achieve this is to construct a committee of classifiers (referred to as an ensemble) to obtain a better predictive model than its constituent models. However, the selection of the best classification methods and the construction of the final ensemble still poses a challenging issue. In this study, we focus and analyze the impact of the choice of classifier types on the accuracy of the classifier ensemble; in addition, we explore the effects of the selection criterion and fusion process on the ensemble accuracy as well. Several fusion scenarios were tested and compared based on a real-world case. Our results show that using an ensemble classification leads to an increase in the accuracy of the classifier models. Consequently, the monitoring and control of the considered real-world case can be improved.