Bayesian Classification Model Research Articles

An Objective Prototype-Based Method for Dual-Polarization Radar Clutter Identification

A prototype-based method is developed to discriminate different types of clutter (ground clutter, sea clutter, and insects) from weather echoes using polarimetric measurements and their textures. This method employs a clustering algorithm to generate data groups from the training dataset, each of which is modeled as a weighted Gaussian distribution called a “prototype.” Two classification algorithms are proposed based on the prototypes, namely maximum prototype likelihood classifier (MPLC) and Bayesian classifier (BC). In the MPLC, the probability of a data point with respect to each prototype is estimated to retrieve the final class label under the maximum likelihood criterion. The BC models the probability density function as a Gaussian mixture composed by the prototypes. The class label is obtained under the maximum a posterior criterion. The two algorithms are applied to S-band dual-polarization CP-2 weather radar data in Southeast Queensland, Australia. The classification results for the test dataset are compared with the NCAR fuzzy-logic particle identification algorithm. Generally good agreement is found for weather echo and ground clutter; however, the confusion matrix indicates that the techniques tend to differ from each other on the recognition of insects.

Bayesian approach to identify spike and sharp waves in EEG data of epilepsy patients

Abstract Electroencephalography (EEG) is the most common test being used to diagnose epilepsy. Most abnormal EEG patterns in epilepsy are interictal epileptiform discharges (IEDs), which consist of spike and sharp waves. These two types of waves can be detected in detail by using the Walsh transformation. In this technique, training data consisting of the original data from EEGs and the results of the first- and second-order Walsh transformation are collected to construct IED profiles. In this paper we propose two Bayesian classification models based on the dependence of the IED profiles. Bayesian classification is applied to classify spike and sharp waves resulting from the Walsh transformation. In our case study, the classification model with dependent features assumption gave better results than the model with independent features assumption.

Distinguishing real from fake ivory products by elemental analyses: A Bayesian hybrid classification method

As laws tighten to limit commercial ivory trading and protect threatened species like whales and elephants, increased sales of fake ivory products have become widespread. This study describes a method, handheld X-ray fluorescence (XRF) as a noninvasive technique for elemental analysis, to differentiate quickly between ivory (Asian and African elephant, mammoth) from non-ivory (bones, teeth, antler, horn, wood, synthetic resin, rock) materials. An equation consisting of 20 elements and light elements from a stepwise discriminant analysis was used to classify samples, followed by Bayesian binary regression to determine the probability of a sample being ‘ivory’, with complementary log log analysis to identify the best fit model for this purpose. This Bayesian hybrid classification model was 93% accurate with 92% precision in discriminating ivory from non-ivory materials. The method was then validated by scanning an additional ivory and non-ivory samples, correctly identifying bone as not ivory with >95% accuracy, except elephant bone, which was 72%. It was less accurate for wood and rock (25–85%); however, a preliminary screening to determine if samples are not Ca-dominant could eliminate inorganic materials. In conclusion, elemental analyses by XRF can be used to identify several forms of fake ivory samples, which could have forensic application.

Automatic Detection Technology of Sonar Image Target Based on the Three-Dimensional Imaging

With 3D imaging of the multisonar beam and serious interference of image noise, detecting objects based only on manual operation is inefficient and also not conducive to data storage and maintenance. In this paper, a set of sonar image automatic detection technologies based on 3D imaging is developed to satisfy the actual requirements in sonar image detection. Firstly, preprocessing was conducted to alleviate the noise and then the approximate position of object was obtained by calculating the signal-to-noise ratio of each target. Secondly, the separation of water bodies and strata is realized by maximum variance between clusters (OTSU) since there exist obvious differences between these two areas. Thus image segmentation can be easily implemented on both. Finally, the feature extraction is carried out, and the multidimensional Bayesian classification model is established to do classification. Experimental results show that the sonar-image-detection technology can effectively detect the target and meet the requirements of practical applications.

In silico target prediction for elucidating the mode of action of herbicides including prospective validation

The rapid emergence of pesticide resistance has given rise to a demand for herbicides with new mode of action (MoA). In the agrochemical sector, with the availability of experimental high throughput screening (HTS) data, it is now possible to utilize in silico target prediction methods in the early discovery phase to suggest the MoA of a compound via data mining of bioactivity data. While having been established in the pharmaceutical context, in the agrochemical area this approach poses rather different challenges, as we have found in this work, partially due to different chemistry, but even more so due to different (usually smaller) amounts of data, and different ways of conducting HTS. With the aim to apply computational methods for facilitating herbicide target identification, 48,000 bioactivity data against 16 herbicide targets were processed to train Laplacian modified Naïve Bayesian (NB) classification models. The herbicide target prediction model (“HerbiMod”) is an ensemble of 16 binary classification models which are evaluated by internal, external and prospective validation sets. In addition to the experimental inactives, 10,000 random agrochemical inactives were included in the training process, which showed to improve the overall balanced accuracy of our models up to 40%. For all the models, performance in terms of balanced accuracy of≥80% was achieved in five-fold cross validation. Ranking target predictions was addressed by means of z-scores which improved predictivity over using raw scores alone. An external testset of 247 compounds from ChEMBL and a prospective testset of 394 compounds from BASF SE tested against five well studied herbicide targets (ACC, ALS, HPPD, PDS and PROTOX) were used for further validation. Only 4% of the compounds in the external testset lied in the applicability domain and extrapolation (and correct prediction) was hence impossible, which on one hand was surprising, and on the other hand illustrated the utilization of using applicability domains in the first place. However, performance better than 60% in balanced accuracy was achieved on the prospective testset, where all the compounds fell within the applicability domain, and which hence underlines the possibility of using target prediction also in the area of agrochemicals.

Predicting Rat and Human Pregnane X Receptor Activators Using Bayesian Classification Models.

The pregnane X receptor (PXR) is a ligand-activated transcription factor that acts as a master regulator of metabolizing enzymes and transporters. To avoid adverse drug-drug interactions and diseases such as steatosis and cancers associated with PXR activation, identifying drugs and chemicals that activate PXR is of crucial importance. In this work, we developed ligand-based predictive computational models for both rat and human PXR activation, which allowed us to identify potentially harmful chemicals and evaluate species-specific effects of a given compound. We utilized a large publicly available data set of nearly 2000 compounds screened in cell-based reporter gene assays to develop Bayesian quantitative structure-activity relationship models using physicochemical properties and structural descriptors. Our analysis showed that PXR activators tend to be hydrophobic and significantly different from nonactivators in terms of their physicochemical properties such as molecular weight, logP, number of rings, and solubility. Our Bayesian models, evaluated by using 5-fold cross-validation, displayed a sensitivity of 75% (76%), specificity of 76% (75%), and accuracy of 89% (89%) for human (rat) PXR activation. We identified structural features shared by rat and human PXR activators as well as those unique to each species. We compared rat in vitro PXR activation data to in vivo data by using DrugMatrix, a large toxicogenomics database with gene expression data obtained from rats after exposure to diverse chemicals. Although in vivo gene expression data pointed to cross-talk between nuclear receptor activators that is captured only by in vivo assays, overall we found broad agreement between in vitro and in vivo PXR activation. Thus, the models developed here serve primarily as efficient initial high-throughput in silico screens of in vitro activity.

Noninvasive Qualitative and Quantitative Assessment of Spoilage Attributes of Chilled Pork Using Hyperspectral Scattering Technique

The objective of this research was to develop a rapid noninvasive method for quantitative and qualitative determination of chilled pork spoilage. Microbiological, physicochemical, and organoleptic characteristics such as the total viable count (TVC), Pseudomonas spp., total volatile basic-nitrogen (TVB-N), pH value, and color parameter L* were determined to appraise pork quality. The hyperspectral scattering characteristics from 54 meat samples were fitted by four-parameter modified Gompertz function accurately. Support vector machines (SVM) was applied to establish quantitative prediction model between scattering fitting parameters and reference values. In addition, partial least squares discriminant analysis (PLS-DA) and Bayesian analysis were utilized as supervised and unsupervised techniques for the qualitative identification of meat spoilage. All stored chilled meat samples were classified into three grades: "fresh," "semi-fresh," and "spoiled." Bayesian classification model was superior to PLS-DA with overall classification accuracy of 92.86%. The results demonstrated that hyperspectral scattering technique combined with SVM and Bayesian possessed a powerful capability for meat spoilage assessment rapidly and noninvasively.

Body size distributions signal a regime shift in a lake ecosystem.

Communities of organisms, from mammals to microorganisms, have discontinuous distributions of body size. This pattern of size structuring is a conservative trait of community organization and is a product of processes that occur at multiple spatial and temporal scales. In this study, we assessed whether body size patterns serve as an indicator of a threshold between alternative regimes. Over the past 7000 years, the biological communities of Foy Lake (Montana, USA) have undergone a major regime shift owing to climate change. We used a palaeoecological record of diatom communities to estimate diatom sizes, and then analysed the discontinuous distribution of organism sizes over time. We used Bayesian classification and regression tree models to determine that all time intervals exhibited aggregations of sizes separated by gaps in the distribution and found a significant change in diatom body size distributions approximately 150 years before the identified ecosystem regime shift. We suggest that discontinuity analysis is a useful addition to the suite of tools for the detection of early warning signals of regime shifts.

Use of handheld X-ray fluorescence as a non-invasive method to distinguish between Asian and African elephant tusks.

We describe the use of handheld X-ray fluorescence, for elephant tusk species identification. Asian (n = 72) and African (n = 85) elephant tusks were scanned and we utilized the species differences in elemental composition to develop a functional model differentiating between species with high precision. Spatially, the majority of measured elements (n = 26) exhibited a homogeneous distribution in cross-section, but a more heterologous pattern in the longitudinal direction. Twenty-one of twenty four elements differed between Asian and African samples. Data were subjected to hierarchical cluster analysis followed by a stepwise discriminant analysis, which identified elements for the functional equation. The best equation consisted of ratios of Si, S, Cl, Ti, Mn, Ag, Sb and W, with Zr as the denominator. Next, Bayesian binary regression model analysis was conducted to predict the probability that a tusk would be of African origin. A cut-off value was established to improve discrimination. This Bayesian hybrid classification model was then validated by scanning an additional 30 Asian and 41 African tusks, which showed high accuracy (94%) and precision (95%) rates. We conclude that handheld XRF is an accurate, non-invasive method to discriminate origin of elephant tusks provides rapid results applicable to use in the field.

A Bayesian Classification Model for Fraud Detection over ATM Platforms

A Bayesian Classification Model for Fraud Detection over ATM Platforms

Web Video Mining: Metadata Predictive Analysis using Classification Techniques

Now a days, the Data Engineering becoming emerging trend to discover knowledge from web audio- visual data such as- YouTube videos, Yahoo Screen, Face Book videos etc. Different categories of web video are being shared on such social websites and are being used by the billions of users all over the world. The uploaded web videos will have different kind of metadata as attribute information of the video data. The metadata attributes defines the contents and features/characteristics of the web videos conceptually. Hence, accomplishing web video mining by extracting features of web videos in terms of metadata is a challenging task. In this work, effective attempts are made to classify and predict the metadata features of web videos such as length of the web videos, number of comments of the web videos, ratings information and view counts of the web videos using data mining algorithms such as Decision tree J48 and navie Bayesian algorithms as a part of web video mining. The results of Decision tree J48 and navie Bayesian classification models are analyzed and compared as a step in the process of knowledge discovery from web videos.

Classification of clinically useful sentences in clinical evidence resources

Most patient care questions raised by clinicians can be answered by online clinical knowledge resources. However, important barriers still challenge the use of these resources at the point of care. ObjectiveTo design and assess a method for extracting clinically useful sentences from synthesized online clinical resources that represent the most clinically useful information for directly answering clinicians’ information needs. Materials and methodsWe developed a Kernel-based Bayesian Network classification model based on different domain-specific feature types extracted from sentences in a gold standard composed of 18 UpToDate documents. These features included UMLS concepts and their semantic groups, semantic predications extracted by SemRep, patient population identified by a pattern-based natural language processing (NLP) algorithm, and cue words extracted by a feature selection technique. Algorithm performance was measured in terms of precision, recall, and F-measure. ResultsThe feature-rich approach yielded an F-measure of 74% versus 37% for a feature co-occurrence method (p<0.001). Excluding predication, population, semantic concept or text-based features reduced the F-measure to 62%, 66%, 58% and 69% respectively (p<0.01). The classifier applied to Medline sentences reached an F-measure of 73%, which is equivalent to the performance of the classifier on UpToDate sentences (p=0.62). ConclusionsThe feature-rich approach significantly outperformed general baseline methods. This approach significantly outperformed classifiers based on a single type of feature. Different types of semantic features provided a unique contribution to overall classification performance. The classifier’s model and features used for UpToDate generalized well to Medline abstracts.

Tracking employment shocks using mobile phone data

Can data from mobile phones be used to observe economic shocks and their consequences at multiple scales? Here we present novel methods to detect mass layoffs, identify individuals affected by them and predict changes in aggregate unemployment rates using call detail records (CDRs) from mobile phones. Using the closure of a large manufacturing plant as a case study, we first describe a structural break model to correctly detect the date of a mass layoff and estimate its size. We then use a Bayesian classification model to identify affected individuals by observing changes in calling behaviour following the plant's closure. For these affected individuals, we observe significant declines in social behaviour and mobility following job loss. Using the features identified at the micro level, we show that the same changes in these calling behaviours, aggregated at the regional level, can improve forecasts of macro unemployment rates. These methods and results highlight promise of new data resources to measure microeconomic behaviour and improve estimates of critical economic indicators.

Fragment virtual screening based on Bayesian categorization for discovering novel VEGFR-2 scaffolds.

The discovery of novel scaffolds against a specific target has long been one of the most significant but challengeable goals in discovering lead compounds. A scaffold that binds in important regions of the active pocket is more favorable as a starting point because scaffolds generally possess greater optimization possibilities. However, due to the lack of sufficient chemical space diversity of the databases and the ineffectiveness of the screening methods, it still remains a great challenge to discover novel active scaffolds. Since the strengths and weaknesses of both fragment-based drug design and traditional virtual screening (VS), we proposed a fragment VS concept based on Bayesian categorization for the discovery of novel scaffolds. This work investigated the proposal through an application on VEGFR-2 target. Firstly, scaffold and structural diversity of chemical space for 10 compound databases were explicitly evaluated. Simultaneously, a robust Bayesian classification model was constructed for screening not only compound databases but also their corresponding fragment databases. Although analysis of the scaffold diversity demonstrated a very unevenly distribution of scaffolds over molecules, results showed that our Bayesian model behaved better in screening fragments than molecules. Through a literature retrospective research, several generated fragments with relatively high Bayesian scores indeed exhibit VEGFR-2 biological activity, which strongly proved the effectiveness of fragment VS based on Bayesian categorization models. This investigation of Bayesian-based fragment VS can further emphasize the necessity for enrichment of compound databases employed in lead discovery by amplifying the diversity of databases with novel structures.

Generation of Bayesian prediction models for OATP-mediated drug–drug interactions based on inhibition screen of OATP1B1, OATP1B1∗15 and OATP1B3

Human organic anion-transporting polypeptide 1B1 (OATP1B1) and OATP1B3 are important hepatic uptake transporters. Early assessment of OATP1B1/1B3-mediated drug–drug interactions (DDIs) is therefore important for successful drug development. A promising approach for early screening and prediction of DDIs is computational modeling. In this study we aimed to generate a rapid, single Bayesian prediction model for OATP1B1, OATP1B1∗15 and OATP1B3 inhibition. Besides our previously generated HEK-OATP1B1 and HEK-OATP1B1∗15 cells, we now generated and characterized HEK-OATP1B3 cells. Using these cell lines we investigated the inhibitory potential of 640 FDA-approved drugs from a commercial library (10μM) on the uptake of [3H]-estradiol-17β-d-glucuronide (1μM) by OATP1B1, OATP1B1∗15, and OATP1B3. Using a cut-off of ⩾60% inhibition, 8% and 7% of the 640 drugs were potent OATP1B1 and OATP1B1∗15 inhibitors, respectively. Only 1% of the tested drugs significantly inhibited OATP1B3, which was not sufficient for Bayesian modeling. Modeling of OATP1B1 and OATP1B1∗15 inhibition revealed that presence of conjugated systems and (hetero)cycles with acceptor/donor atoms in- or outside the ring enhance the probability of a molecule binding these transporters. The overall performance of the model for OATP1B1 and OATP1B1∗15 was ⩾80%, including evaluation with a true external test set. Our Bayesian classification model thus represents a fast, inexpensive and robust means of assessing potential binding of new chemical entities to OATP1B1 and OATP1B1∗15. As such, this model may be used to rank compounds early in the drug development process, helping to avoid adverse effects in a later stage due to inhibition of OATP1B1 and/or OATP1B1∗15.

A Research Using Correlation Coefficient to Make Bayesian Classification Data Mining

In traditional Bayesian classification data mining methods, there may be defects such as predictions unreliable because the selected predictors are little or not related with the target factor. this paper analyzes the correlation between predictors and the target factor using correlation coefficient based on Bayesian classification model and combines with Hadoop distributed file system and parallel programming models to explore an improved algorithm. The experiments show that this method not only makes the prediction more reliable but also saves resources and improves the efficiency of the algorithm greatly. In addition, it is suitable for massive data processing.

Introduction of a methodology for visualization and graphical interpretation of Bayesian classification models.

Supervised machine learning models are widely used in chemoinformatics, especially for the prediction of new active compounds or targets of known actives. Bayesian classification methods are among the most popular machine learning approaches for the prediction of activity from chemical structure. Much work has focused on predicting structure-activity relationships (SARs) on the basis of experimental training data. By contrast, only a few efforts have thus far been made to rationalize the performance of Bayesian or other supervised machine learning models and better understand why they might succeed or fail. In this study, we introduce an intuitive approach for the visualization and graphical interpretation of naïve Bayesian classification models. Parameters derived during supervised learning are visualized and interactively analyzed to gain insights into model performance and identify features that determine predictions. The methodology is introduced in detail and applied to assess Bayesian modeling efforts and predictions on compound data sets of varying structural complexity. Different classification models and features determining their performance are characterized in detail. A prototypic implementation of the approach is provided.

Visualization and Graphical Interpretation of Bayesian Compound Classification Models

Visualization and Graphical Interpretation of Bayesian Compound Classification Models

Novel Bayesian classification models for predicting compounds blocking hERG potassium channels.

A large number of drug-induced long QT syndromes are ascribed to blockage of hERG potassium channels. The aim of this study was to construct novel computational models to predict compounds blocking hERG channels. Doddareddy's hERG blockage data containing 2644 compounds were used, which divided into training (2389) and test (255) sets. Laplacian-corrected Bayesian classification models were constructed using Discovery Studio. The models were internally validated with the training set of compounds, and then applied to the test set for validation. Doddareddy's experimentally validated dataset with 60 compounds was used for external test set validation. A Bayesian classification model considering the effects of four molecular properties (Mw, PPSA, ALogP and pKa_basic) as well as extended-connectivity fingerprints (ECFP_14) exhibited a global accuracy (91%), parameter sensitivity (90%) and specificity (92%) in the test set validation, and a global accuracy (58%), parameter sensitivity (61%) and specificity (57%) in the external test set validation. The novel model is better than those in the literatures for predicting compounds blocking hERG channels, and can be used for large-scale prediction.

ANN and Bayesian Classification Models for Virtual Screening of Endocrine-Disrupting Chemicals

The identification of endocrine-disrupting chemicals (EDCs) is one of the important goals of environmental chemical hazard screening. The adverse health effects of EDCs in humans have been demonstrated to involve the developmental, reproductive, neurological, cardiovascular, metabolic, and immune systems. The present study reports QSAR classification studies on a large database comprising 8,212 compounds collected from the Estrogenic Activity Database and National Center for Biotechnology Information Database. In this study, four classification models (Bayesian Categorization Model with molecular fingerprints or molecular descriptors as an input and Neural Classification Models with and without Bayesian regularization) were used. Evaluation of these binomial classification methods indicated that the Bayesian method (Bayesian QSAR) works as an excellent method for prediction with fingerprints used as input. In the case of the multilayer perceptron with molecular descriptors as inputs, changing the training mode by introducing a Bayesian regularization algorithm significantly improved ANNs' predictive power. Our goal was to test two popular classification methods suitable for processing large data sets. Such datasets were required to ensure the prediction performances and applicability of the models as a virtual screening tool for an extensive database.