Bayesian Classification Approach Research Articles

Federated Bayesian network approach for cross-regional air pollution classification: a case study of the Beijing-Tianjin-Hebei region.

Although machine learning methods have enabled considerable progress in air quality assessment, challenges persist regarding data privacy, cross-regional data processing, and model generalization. To address these issues, we introduce an advanced federated Bayesian network (FBN) approach. By integrating federated learning, adaptive optimization algorithms, and homomorphic encryption technologies, we substantially enhanced the efficiency and security of cross-regional air quality data processing. The novelty of this research lies in the improvements implemented in federated learning for air quality data analysis, particularly in distributed model training optimization and data consistency. Through the integration of adaptive structural modification strategies and simulated annealing immune optimization algorithms, we markedly enhanced the structural learning accuracy of the Bayesian network, resulting in a 20% improvement in prediction accuracy. Moreover, employing homomorphic encryption ensured data transmission security and confidentiality. In our Beijing-Tianjin-Hebei case study, our method demonstrated a 15% improvement in air quality classification accuracy compared to conventional methods and exhibited superior interpretability in analyzing environmental factor interactions. We quantified complex air pollution patterns across regions and found that a 30% fluctuation in the air quality index correlated with NO2 concentrations. We also observed a moderate positive correlation between specific pollutant indicators in Hebei Province and Tianjin and changes in air quality. Additionally, the FBN exhibited better operational efficiency and data confidentiality than other machine learning models in handling large-scale and multisource environmental data. Our FBN approach presents a novel perspective for environmental monitoring and assessment, vital for understanding complex air pollution patterns and formulating future ecological protection policies.

Disentangling magnitude processing, natural number biases, and benchmarking in fraction comparison tasks: A person-centered Bayesian classification approach

Research on fraction comparison shows that students often follow biased comparison strategies, in particular such strategies that build on their knowledge of natural numbers. On the other hand they also apply successful comparison strategies such as benchmarking or fraction magnitude processing. Which strategies are applied or even combined depends on the students’ knowledge and on the task type. To investigate these complex relationships, we developed a balanced 2 × 2-dimensional itemset (congruent vs. incongruent items; benchmarking vs. non-benchmarking items) and a Bayesian classification of individual students’ performance (solution patters, response time, and individual distance effect), which we applied to an assessment of N = 350 sixth graders. We could show that the classification of the students with respect to possible solution strategies matched our hypotheses: We could replicate existing patterns and found additional composite strategies such as ‘benchmarking or bias‘ with a bias only in solution rates of non-benchmark items. In further analyses we found ‘benchmarking or suppressed bias-strategies (i.e., a bias in problem solving time of non-benchmarking items). Our study extends previous knowledge on individual strategies in fraction comparison and proposes a new person-centered approach to classify individual student profiles even with small profile sizes.

Naïve Bayesian Classification Based Glioma Brain Tumor Segmentation Using Grey Level Co-occurrence Matrix Method

Brain tumors vary widely in size and form, making detection and diagnosis difficult. This study's main aim is to identify abnormal brain images., classify them from normal brain images, and then segment the tumor areas from the categorised brain images. In this study, we offer a technique based on the Nave Bayesian classification approach that can efficiently identify and segment brain tumors. Noises are identified and filtered out during the preprocessing phase of tumor identification. After preprocessing the brain image, GLCM and probabilistic properties are extracted. Naive Bayesian classifier is then used to train and label the retrieved features. When the tumors in a brain picture have been categorised, the watershed segmentation approach is used to isolate the tumors. This paper's brain pictures are from the BRATS 2015 data collection. The suggested approach has a classification rate of 99.2% for MR pictures of normal brain tissue and a rate of 97.3% for MR images of aberrant Glioma brain tissue. In this study, we provide a strategy for detecting and segmenting tumors that has a 97.54% Probability of Detection (POD), a 92.18% Probability of False Detection (POFD), a 98.17% Critical Success Index (CSI), and a 98.55% Percentage of Corrects (PC). The recommended Glioma brain tumour detection technique outperforms existing state-of-the-art approaches in POD, POFD, CSI, and PC because it can identify tumour locations in abnormal brain images.

Bayesian feature selection for radiomics using reliability metrics.

Introduction: Imaging of tumors is a standard step in diagnosing cancer and making subsequent treatment decisions. The field of radiomics aims to develop imaging based biomarkers using methods rooted in artificial intelligence applied to medical imaging. However, a challenging aspect of developing predictive models for clinical use is that many quantitative features derived from image data exhibit instability or lack of reproducibility across different imaging systems or image-processing pipelines. Methods: To address this challenge, we propose a Bayesian sparse modeling approach for image classification based on radiomic features, where the inclusion of more reliable features is favored via a probit prior formulation. Results: We verify through simulation studies that this approach can improve feature selection and prediction given correct prior information. Finally, we illustrate the method with an application to the classification of head and neck cancer patients by human papillomavirus status, using as our prior information a reliability metric quantifying feature stability across different imaging systems.

A Bayesian optimal convolutional neural network approach for classification of coal and gangue with multispectral imaging

The precise classification of coal and gangue is a crucial link for effective sorting and efficient utilization. However, there are some shortcomings in traditional methods, such as water consumption, coal slime pollution, and great influence of environmental factors, and so on. Here, multispectral imaging technology combined with the convolutional neural network (CNN) was applied to classify coal and gangue, in which the hyperparameters of the CNN model were optimized by Bayesian algorithm. The multispectral images in the range of 675–975 nm of 209 pieces of coal and 201 pieces of gangue, which came from the Huainan mining area, were collected. The CNN and traditional modeling methods (combination strategy of image feature extraction and classifier) were employed to develop identification models, and the classification results were analyzed and compared on the multispectral dataset of coal and gangue. The identification analysis model based on CNN had the best performance, and the F1 score reached 1.00. At this time, the hyperparameters of the model are as follows: network depth was 1, initial learning rate was 0.012939, random gradient descent momentum was 0.83813, and L2 regularization intensity was 0.0099852. Moreover, the robustness of the CNN identification model was verified by introducing different levels of noise signals. The identification analysis model based on the CNN can quickly and accurately identify coal and gangue without complex image processing steps, and the model has certain anti-interference ability, which will promote the progress of automatic separation technology for coal and gangue.

Application of PZT Ceramic Sensors for Composite Structure Monitoring Using Harmonic Excitation Signals and Bayesian Classification Approach.

The capabilities of ceramic PZT transducers, allowing for elastic wave excitation in a broad frequency spectrum, made them particularly suitable for the Structural Health Monitoring field. In this paper, the approach to detecting impact damage in composite structures based on harmonic excitation of PZT sensor in the so-called pitch–catch PZT network setup is studied. In particular, the repeatability of damage indication for similar configuration of two independent PZT networks is analyzed, and the possibility of damage indication for different localization of sensing paths between pairs of PZT sensors with respect to damage locations is investigated. The approach allowed for differentiation between paths sensitive to the transmission mode of elastic wave interaction and sensitive reflection mode. In addition, a new universal Bayesian approach to SHM data classification is provided in the paper. The defined Bayesian classifier is based on asymptotic properties of Maximum Likelihood estimators and Principal Component Analysis for orthogonal data transformation. Properties of the defined algorithm are compared to the standard nearest-neighbor classifier based on the acquired experimental data. It was shown in the paper that the proposed approach is characterized by lower false-positive indications in comparison with the nearest-neighbor algorithm.

Reservoir characterization over the Lille Prinsen and Ivar Aasen fields in the Norwegian North Sea using ocean-bottom-node seismic data — A case study

We have developed an integrated workflow for estimating elastic parameters within the Late Triassic Skagerrak Formation, the Middle Jurassic Sleipner and Hugin Formations, the Paleocene Heimdal Formation, and the Eocene Grid Formation in the Utsira High area of the Norwegian North Sea. Our workflow begins with petrophysical analysis carried out at the available wells. Then, model-based prestack simultaneous impedance inversion outputs were derived, and attempts were made to estimate the petrophysical parameters (the volume of shale, porosity, and water saturation) from seismic data using extended elastic impedance. On not obtaining convincing results, we switched over to multiattribute regression analysis for estimating them, which yielded encouraging results. Finally, the Bayesian classification approach was used for defining different facies in the intervals of interest.

Efficient posterior sampling for high-dimensional imbalanced logistic regression

Classification with high-dimensional data is of widespread interest and often involves dealing with imbalanced data. Bayesian classification approaches are hampered by the fact that current Markov chain Monte Carlo algorithms for posterior computation become inefficient as the number [Formula: see text] of predictors or the number [Formula: see text] of subjects to classify gets large, because of the increasing computational time per step and worsening mixing rates. One strategy is to employ a gradient-based sampler to improve mixing while using data subsamples to reduce the per-step computational complexity. However, the usual subsampling breaks down when applied to imbalanced data. Instead, we generalize piecewise-deterministic Markov chain Monte Carlo algorithms to include importance-weighted and mini-batch subsampling. These maintain the correct stationary distribution with arbitrarily small subsamples and substantially outperform current competitors. We provide theoretical support for the proposed approach and demonstrate its performance gains in simulated data examples and an application to cancer data.

A new complement naïve Bayesian approach for biomedical data classification

Biomedical data classification tasks are very challenging because data is usually large, noised and imbalanced. Particularly the noise can reduce system performance in terms of classification accuracy, time in building a classifier and the size of the classifier. Accordingly, most existing learning algorithms have integrated various approaches to enhance their learning abilities from noisy environments, but the existence of noise can still introduce serious negative impacts. A more reasonable solution might be to employ some preprocessing mechanisms to handle noisy instances before a learner is formed. Therefore, we introduce a method called double learning to improve the classification performance of our model. As to the author’s knowledge, most of the previous works used the normal (noise free) instances for model construction (training) after the noise instances are isolated. This approach increases computational task on model construction for active learners and total computational time for passive learners. It also ignores minority data instance which leads to miss classification of instances from minority group as test cases. The main idea of this paper is to construct a model using noised instances. This approach minimizes the model construction time by reducing the number of instances and improves classification performance. Therefore, only the identified noised data are used for model construction instead of the normal (noise free) data. Since noised instances are used for model construction, the entire naive Bayesian working logic is reversed. This method is called complement naive Bayesian (CNB) which makes use of the idea of complement based learning to improve the accuracy performance. Finally, the performance of the proposed CNB is compared to naive Bayesian and some other classification algorithms with the single photon emission computed tomography, Indian liver patient dataset, Wilt and Tic-Tac-Toe endgame datasets. The experimental results demonstrated that the proposed approach has shown promising results in terms of computational time and accuracy performance on both balanced and imbalanced datasets used.

Unsupervised Bayesian Classification of a Hyperspectral Image Based on the Spectral Mixture Model and Markov Random Field

Typical unsupervised classification of hyperspectral imagery (HSI) uses a Gaussian mixture model to determine intensity similarity of pixels. However, the existence of mixed pixels in HSI tends to reduce the effectiveness of the similarity measure and leads to large classification errors. Since a semantic class is always dominated by a particular endmember, a mixed pixel can be better classified by identifying the dominant endmember. By exploiting the spectral mixture model (SMM) that describes the endmember-abundance pattern of mixed pixels, the discriminative ability of HSI can be enhanced. A Bayesian classification approach is presented for spatial–spectral HSI classification, where the data likelihood is built upon the SMM, and the label prior is based on a Markov random field (MRF). The new approach has three key characteristics. First, instead of using intensity similarity, the new approach uses the abundance-endmember pattern of each pixel and classifies a pixel by its dominant endmember. Second, to integrate the SMM into a Bayesian framework, a data likelihood is designed based on the SMM to reflect the influence of the dominant endmember on the conditional distribution of the mixed pixel given the class label. Third, the resulting maximum a posteriori problem is solved by the expectation–maximization (EM) algorithm, in which the E-step adopts a graph-cut approach to estimate the class labels, and the M-step adopts a purified-means approach to estimate the endmembers. Experiments on both simulated and real HSIs demonstrate that the proposed method can exploit the spatial–spectral information of HSI to achieve high accuracy in unsupervised classification of HSI.

A Bayesian classification approach for predicting Gesonia gemma Swinhoe population on soybean crop in relation to abiotic factors based on economic threshold level

Predicting of insect pest population with accuracy and speed when given large data set will make a major contribution to the success of integrated pest management. Naive Bayesian classification has been proposed for predicting the insect pest Gesonia gemma Swinhoe on soybean crop. The Naive Bayesian classifier works based on Bayes’ theorem and can predict class probabilities that a given tuple from the dataset belongs to a particular class. The dataset includes abiotic factors as features along with the class feature (pest incidence) are separated as training data and testing data, then the model was built on the training set by finding the probability for each of its features in relation with the class feature. The Naive Bayesian classification from the trained model, best fits the testing data with 90% accuracy, thus the proposed approach can be very useful in predicting the pest G. gemma on soybean crop.

Discovery of VEGFR2 inhibitors by integrating naïve Bayesian classification, molecular docking and drug screening approaches.

The high morbidity and mortality of cancer make it one of the leading causes of global death, thus it is an urgent need to develop effective drugs for cancer therapy. Vascular endothelial growth factor receptor-2 (VEGFR2) acts as a central modulator of angiogenesis, and is therefore an important pharmaceutical target for developing anti-angiogenic agents. In this study, ligand-based naïve Bayesian (NB) models and structure-based molecular docking were combined to develop a virtual screening (VS) pipeline for identifying potential VEGFR2 inhibitors from FDA-approved drugs. The best validated naïve Bayesian model (NB-c) gave Matthews correlation coefficients of 0.966 and 0.951 for the test set and external validation set, respectively. 1841 FDA-approved drugs were sequentially screened by the optimal model NB-c and molecular docking module LibDock. By analyzing the results of VS, 9 top ranked drugs with EstPGood value ≥ 0.6 and LibDock Score ≥ 120 were chosen for biological validation. VEGFR2 kinase assay results demonstrated that flubendazole, rilpivirine and papaverine showed VEGFR2 inhibitory activities with IC50 values ranging from 0.47 to 6.29 μM. Binding mode analysis with CDOCKER revealed the action mechanism of the 3 hit drugs binding to VEGFR2. In summary, we not only proposed an integrated VS pipeline for potential VEGFR2 inhibitors screening, but also identified 3 FDA-approved drugs as novel VEGFR2 inhibitors, which could be used to design and develop new antiangiogenic agents.

On-line crop/weed discrimination through the Mahalanobis distance from images in maize fields

This study proposes a new automatic method for crop/weed discrimination in images captured in maize fields during the initial growth stages. The images were obtained under perspective projection with a camera installed on board at the front part of a tractor. Different approaches have addressed the problem based on crop row determination and then assuming that inter-row plants are weeds. Nevertheless, an important challenge is the identification of weeds intermixed within the crop rows. This issue is addressed on this paper by applying a minimum criterion distance based on the Mahalanobis distance derived from a Bayesian classification approach, this makes the main contribution. The identification of both intra- and inter-row weeds is useful for more accurate weed quantification for site-specific treatments. Image quality is affected by uncontrolled lighting conditions in outdoor agricultural environments. Also, different plant densities appear due to different growth stages affecting the crop/weed identification process. The proposed method was designed to deal with the above undesired situations, consisting of three phases: (i) segmentation, (ii) training and (iii) testing. The three phases are executed on-line for each image, where training is specific of each single image, requiring no prior training, as it is usual in common machine learning-based approaches, mainly supervised. This makes the second research contribution. The performance of the proposed approach was quantitatively compared against three existing strategies, achieving an accuracy of 91.8%, pixel-wise determined against ground-truth images manually built, with processing times ≤280 ms, which can be useful for real-time applications.

Detection and Attribution of Multivariate Climate Change Signals Using Discriminant Analysis and Bayesian Theorem

Abstract Detection and attribution methods in climatological research aim at assessing whether observed climate anomalies and trends are still consistent with the range of natural climate variations or rather an indication of anthropogenic climate change. In this study, the authors pursue a novel approach by using discriminant analysis to enhance the distinction between past and future climates from state-of-the-art climate model simulations. The method is based on multivariate fingerprints that are defined in the space of several prominent climate indices representing the thermal, dynamical, and hygric aspects of climate change. Attribution is carried out by means of a Bayesian classification approach. The leading discriminant function accounts for more than 99% of total discriminability, with temperature variables, extratropical precipitation, and extratropical circulation modes mainly contributing to the discriminant power. The misclassification probability between probability density functions of past and future climates is substantially reduced by the discriminant analysis: from >50% to <15%. Since the mid-1980s, the observed anomalies of the considered climate indices are more or less consistently attributed to a climate under strong radiative forcing, projected for the first half of the twenty-first century. The authors also assess the sensitivity of their results to different emissions scenarios from the CMIP3 and CMIP5 multimodel ensembles, seasons, prior probabilities for the early twenty-first-century climate, estimates of the observational error, low-pass filters, variable compositions, group numbers, and reference data.

Prediction of chemical biodegradability using computational methods

Biodegradability is a key factor to describe the long-time effects of chemicals to be decomposed in the environment. Compared with time-consuming and laborious experimental testing, the use of in silico approaches for assessing chemical biodegradability is highly encouraged by the legislators. In this study, based on an extensive data-set with 547 ready biodegradation (RB) and 1178 non-ready biodegradation (NRB) chemicals, we first examined the differences of the important physico-chemical properties and scaffold architectures between the RB and NRB molecules. We found that compared with the NRB molecules, the RB molecules are usually smaller, more flexible and hydrophilic, and have less polar groups and more complicated structural patterns (ring systems). However, the RB and NRB molecules cannot be well distinguished by any simple property-based or substructure-based rules. Then, the naïve Bayesian classification (NBC) approach was employed to develop classifiers for discriminating the RB and NRB molecules. Based on the 21 physico-chemical properties, 76 VolSurf descriptors and LPFP_4 structural fingerprints, the Bayesian classifier can achieve a sensitivity of 0.877, a specificity of 0.864, a global accuracy of 0.869, a C value of 0.720 and a AUC value of 0.890 for the training set. Besides, the best predictions can be achieved for the classifiers based on the combinations of simple physico-chemical properties, VolSurf descriptors, and LPFP_6 fingerprints for the test set I (AUC = 0.921), and any of the three fingerprint classes (ECFC_6, ECFC_8 or LPFC_4) for the test set II (AUC = 0.901). In addition, 20 structural fragments favourable and unfavourable for ready biodegradation, which were directly generated from the best naive Bayesian classifier, were highlighted and discussed. The results provide useful guidelines/tools for designing promising chemical compounds with good chemical biodegradability.

A Bayesian state-space approach for damage detection and classification

The problem of automatic damage detection in civil structures is complex and requires a system that can interpret collected sensor data into meaningful information. We apply our recently developed switching Bayesian model for dependency analysis to the problems of damage detection and classification. The model relies on a state-space approach that accounts for noisy measurement processes and missing data, which also infers the statistical temporal dependency between measurement locations signifying the potential flow of information within the structure. A Gibbs sampling algorithm is used to simultaneously infer the latent states, parameters of the state dynamics, the dependence graph, and any changes in behavior. By employing a fully Bayesian approach, we are able to characterize uncertainty in these variables via their posterior distribution and provide probabilistic estimates of the occurrence of damage or a specific damage scenario. We also implement a single class classification method which is more realistic for most real world situations where training data for a damaged structure is not available. We demonstrate the methodology with experimental test data from a laboratory model structure and accelerometer data from a real world structure during different environmental and excitation conditions.

A Bayesian approach for probabilistic classification and inference of occupant thermal preferences in office buildings

This paper presents a new data-driven method for learning personalized thermal preference profiles, by formulating a combined classification and inference problem, without developing different models for each occupant. Different from existing approaches, we developed a generalized thermal preference model in which our main hypothesis, “Different people prefer different thermal conditions”, is explicitly encoded. The approach is fully Bayesian, and it is based on the premise that the thermal preference is mainly governed by (i) an overall thermal stress, represented using physical process equations with relatively few parameters along with prior knowledge of the parameters, and (ii) the personal thermal preference characteristic, which is modeled as a hidden random variable. The concept of clustering occupants based on this hidden variable, i.e., similar thermal preference characteristic, is introduced. The results, based on a dataset collected from a typical office building population, show clear evidence of the existence of multi-clusters; in particular, the 5-cluster model performed best compared to 2, 3 and higher cluster models using the studied dataset. Subsequently, the thermal preference of a new occupant in the dataset is inferred by using a mixture of the general sub-models for each cluster. The results show that the method developed in this study provides accurate predictions for personalized thermal preference profiles and it is efficient as it only requires a relatively small dataset collected from each occupant. The approach presented in this paper is a significant step towards personalized environments in office buildings using real-time feedback from occupants.

Detection of the Onset of Diabetes Mellitus by Bayesian Classifier Based Medical Expert System

Expert systems play an important role in medical diagnosis research. Researches are still being conducted for building expert systems capable of diagnosing different diseases. Diabetes mellitus is one of the diseases that have gained attention in the past years. Patients are usually unaware of having this disease and are finally diagnosed with diabetes after several years from onset. Since diabetes can be controlled, it is much desirable to harness it at the onset. Therefore, the prediction of onset of diseases like diabetes has been the point of interest for the researchers. Researchers are continuously trying to formulate an inference engine, a part of an expert system, in order to predict the disease at the beginning. In this paper, we present a Bayesian classification approach to identify the onset of diabetes mellitus in patients using a well-known data set as the sample. We have found an intriguing result with more than 87% accuracy.

ADMET Evaluation in Drug Discovery. 16. Predicting hERG Blockers by Combining Multiple Pharmacophores and Machine Learning Approaches.

Blockade of human ether-à-go-go related gene (hERG) channel by compounds may lead to drug-induced QT prolongation, arrhythmia, and Torsades de Pointes (TdP), and therefore reliable prediction of hERG liability in the early stages of drug design is quite important to reduce the risk of cardiotoxicity-related attritions in the later development stages. In this study, pharmacophore modeling and machine learning approaches were combined to construct classification models to distinguish hERG active from inactive compounds based on a diverse data set. First, an optimal ensemble of pharmacophore hypotheses that had good capability to differentiate hERG active from inactive compounds was identified by the recursive partitioning (RP) approach. Then, the naive Bayesian classification (NBC) and support vector machine (SVM) approaches were employed to construct classification models by integrating multiple important pharmacophore hypotheses. The integrated classification models showed improved predictive capability over any single pharmacophore hypothesis, suggesting that the broad binding polyspecificity of hERG can only be well characterized by multiple pharmacophores. The best SVM model achieved the prediction accuracies of 84.7% for the training set and 82.1% for the external test set. Notably, the accuracies for the hERG blockers and nonblockers in the test set reached 83.6% and 78.2%, respectively. Analysis of significant pharmacophores helps to understand the multimechanisms of action of hERG blockers. We believe that the combination of pharmacophore modeling and SVM is a powerful strategy to develop reliable theoretical models for the prediction of potential hERG liability.

A Bayesian Classification Approach Using Class-Specific Features for Text Categorization

In this paper, we present a Bayesian classification approach for automatic text categorization using class-specific features. Unlike conventional text categorization approaches, our proposed method selects a specific feature subset for each class. To apply these class-specific features for classification, we follow Baggenstoss's PDF Projection Theorem (PPT) to reconstruct the PDFs in raw data space from the class-specific PDFs in low-dimensional feature subspace, and build a Bayesian classification rule. One noticeable significance of our approach is that most feature selection criteria, such as Information Gain (IG) and Maximum Discrimination (MD), can be easily incorporated into our approach. We evaluate our method's classification performance on several real-world benchmarks, compared with the state-of-the-art feature selection approaches. The superior results demonstrate the effectiveness of the proposed approach and further indicate its wide potential applications in data mining.