Conditional Random Fields Method Research Articles

Probabilistic analysis of lateral behaviour of offshore wind turbine monopile considering uncertainties of geological model and loads

Due to the complex geological, meteorological and hydrological conditions, along with the limited availability of field measurements, multiple uncertainties are related to the performance analysis of the offshore wind turbine. The main sources of uncertainty are the geological model and the environmental loads. However, the geological model uncertainty and loads uncertainty were often characterized separately in previous studies. To this end, this article proposes a novel framework for the coupled characterization of the geological model and loads uncertainties. With this new framework, the geological model is generated using a conditional random field method, in which the stratigraphic and geo-properties uncertainties are characterized simultaneously. Next, environmental parameters are considered as random variables, and the cross correlation between different environmental parameters is captured with the Pearson correlation coefficient. Then, multiple realizations of the geological model and these of environmental parameters are sampled with Monte Carlo simulation. With these generated realizations as inputs, the monopile foundation performance could then be analyzed probabilistically considering uncertainties of the geological model and loads. To demonstrate the effectiveness of the new framework, a case study of an offshore site in Taiwan is conducted. Further, a comparative study is performed to quantitatively reveal the influence of different uncertainty sources on site characterization and monopile behaviour assessment, through which the superiority of the presented framework is demonstrated.

Refined glacial lake extraction in a high-Asia region by deep neural network and superpixel-based conditional random field methods

Abstract. Remote sensing extraction of glacial lakes is an effective way of monitoring water body distribution and outburst events. At present, the lack of glacial lake datasets and the edge recognition problem of semantic segmentation networks lead to poor accuracy and inaccurate outlines of glacial lakes. Therefore, this study constructed a high-resolution dataset containing seven types of glacial lakes and proposed a refined glacial lake extraction method, which combines the LinkNet50 network for rough extraction and simple linear iterative clustering (SLIC) dense conditional random field (DenseCRF) for optimization. The results show that (1) with Google Earth images of 0.52 m resolution in the study area, the recall, precision, F1 score, and intersection over union (IoU) of glacial lake extraction based on the proposed method are 96.52 %, 92.49 %, 94.46 %, and 90.69 %, respectively, and (2) with the Google Earth images of 2.11 m resolution in the Qomolangma National Nature Reserve, 2300 glacial lakes with a total area of 65.17 km2 were detected by the proposed method. The area of the minimum glacial lake that can be extracted is 160 m2 (less than 6×6 pixels). This method has advantages in small glacial lake extraction and refined outline detection, which can be applied to extracting glacial lakes in the high-Asia region with high-resolution images.

A hybrid frequency based, Syntax, and Conditional Random Field Method for Implicit and Explicit Aspect Extraction

A hybrid frequency based, Syntax, and Conditional Random Field Method for Implicit and Explicit Aspect Extraction

Improving Indonesian Named Entity Recognition for Domain Zakat Using Conditional Random Fields

In Indonesia, where the majority of the population is Muslim, one of the obligations of a Muslim is zakat. To reduce illiteracy about zakat among Muslims, they need to have access to basic information about it. In order to facilitate the acquisition of this information, this study utilized named entity recognition (NER) and defined 12 named entity classes for the zakat domain, including the pillars of Islam, various types of zakat, and zakat management institutions. The Conditional Random Fields method was used for testing Indonesian-NER in three scenarios. In the specific context of the Zakat domain, NER can extract information about organizations, individuals, and locations involved in collecting and distributing Zakat funds. This information can improve the Zakat system’s efficiency and transparency and support research and analysis on Zakat-related topics. The average performance evaluation of the Indonesian-NER model showed a precision of 0.902, recall of 0.834, and an F1-score of 0.867.

An Innovative Model of Higher Mathematics Curriculum Education Incorporating Artificial Intelligence Technology

Abstract This paper first proposes the direction of constructing a higher mathematics teaching mode supported by intelligent technology and then models the learner portrait of the learning outcome data according to the subject knowledge graph. An improved ant colony optimization algorithm is used to search for the optimal learning path, which is then combined with an improved convolutional neural network to generate a personalized learning path. The Trans R method is used to quantify the relationship between learners and learning resources and a semi-supervised learning conditional random field method based on K-NN is proposed to label learning resources and generate learning accurate evaluation for smart teaching. The smart teaching model of advanced mathematics courses is applied and analyzed in terms of students’ advanced mathematics pre and post-test scores, students’ satisfaction, teachers’ teaching methods and teaching resources in four directions. The analysis obtained that the posttest scores of the learners in the experimental group were 75.631, and the posttest scores of the learners in the control group were 66.314, with a difference of 9.317. The significance level of the variance chi-square test was 0.000<0.05, which shows that there is a significant difference between the posttest scores of the experimental group and those of the control group, and it indicates that the wisdom teaching of higher mathematics significantly enhances the learning performance of the learners.

An Efficient Method for Deidentifying Protected Health Information in Chinese Electronic Health Records: Algorithm Development and Validation.

BackgroundWith the popularization of electronic health records in China, the utilization of digitalized data has great potential for the development of real-world medical research. However, the data usually contains a great deal of protected health information and the direct usage of this data may cause privacy issues. The task of deidentifying protected health information in electronic health records can be regarded as a named entity recognition problem. Existing rule-based, machine learning–based, or deep learning–based methods have been proposed to solve this problem. However, these methods still face the difficulties of insufficient Chinese electronic health record data and the complex features of the Chinese language.ObjectiveThis paper proposes a method to overcome the difficulties of overfitting and a lack of training data for deep neural networks to enable Chinese protected health information deidentification.MethodsWe propose a new model that merges TinyBERT (bidirectional encoder representations from transformers) as a text feature extraction module and the conditional random field method as a prediction module for deidentifying protected health information in Chinese medical electronic health records. In addition, a hybrid data augmentation method that integrates a sentence generation strategy and a mention-replacement strategy is proposed for overcoming insufficient Chinese electronic health records.ResultsWe compare our method with 5 baseline methods that utilize different BERT models as their feature extraction modules. Experimental results on the Chinese electronic health records that we collected demonstrate that our method had better performance (microprecision: 98.7%, microrecall: 99.13%, and micro-F1 score: 98.91%) and higher efficiency (40% faster) than all the BERT-based baseline methods.ConclusionsCompared to baseline methods, the efficiency advantage of TinyBERT on our proposed augmented data set was kept while the performance improved for the task of Chinese protected health information deidentification.

An Efficient Plant Disease Recognition System Using Hybrid Convolutional Neural Networks (CNNs) and Conditional Random Fields (CRFs) for Smart IoT Applications in Agriculture

In recent times, the Internet of Things (IoT) and Deep Learning Models (DLMs) can be utilized for developing smart agriculture to determine the exact location of the diseased part of the leaf on farmland in an efficient manner. There is no exception that convolutional neural networks (CNNs) have achieved the latest accomplishment in many aspects of human life and the farming sector. Semantic image segmentation is considered the main problem in computer vision. Despite tremendous progress in applications, approximately all semantic image segmentation algorithms fail to achieve sufficient hash results because of the absence of details sensitivity, problems in assessing the global similarity of image pixels, or both. Methods of post-processing improvement, as a wonderfully critical means of improving the underlying flaws mentioned above from algorithms, depend almost on Conditional Random Fields (CRFs). Therefore, plant disease prediction plays important role in the premature notification of the disease to alleviate its effects on disease forecast investigation purposes in the smart farming arena. Hence, this work proposes an efficient IoT-based plant disease recognition system using semantic segmentation methods such as FCN-8 s, CED-Net, SegNet, DeepLabv3, and U-Net with the CRF method to allocate disease parts in leaf crops. Evaluation of this network and comparison with other networks of the state art. The experimental results and their comparisons proclaim over F1-score, sensitivity, and intersection over union (IoU). The proposed system with SegNet and CRFs gives high results compared with other methods. The superiority and effectiveness of the mentioned improvement method, as well as its range of implementation, are confirmed through experiments.

A conditional random field based feature learning framework for battery capacity prediction

This paper proposes a network model framework based on long and short-term memory (LSTM) and conditional random field (CRF) to promote Li-ion battery capacity prediction results. The model uses LSTM to extract temporal features from the data and CRF to build a transfer matrix to enhance temporal feature learning for long serialization prediction of lithium battery feature sequence data. The NASA PCOE lithium battery dataset is selected for the experiments, and control tests on LSTM temporal feature extraction modules, including recurrent neural network (RNN), gated recurrent unit (GRU), bi-directional gated recurrent unit (BiGRU) and bi-directional long and short term memory (BiLSTM) networks, are designed to test the adaptability of the CRF method to different temporal feature extraction modules. Compared with previous Li-ion battery capacity prediction methods, the network model framework proposed in this paper achieves better prediction results in terms of root mean square error (RMSE) and mean absolute percentage error (MAPE) metrics.

Unsupervised SAR Image Change Detection for Few Changed Area Based on Histogram Fitting Error Minimization

Change detection in synthetic aperture radar (SAR) images is an essential task of remote sensing image analysis. However, the thresholding procedure is the main difficulty in change detection for few changed areas for traditional change detection methods. In this paper, we propose a novel change detection method for very few changed or even none changed areas. The proposed method contains three procedures: difference image (DI) generation, thresholding, and spatial analysis. In the second procedure, a new thresholding method called Histogram Fitting Error Minimization (HFEM) is proposed for few changed areas. HFEM is derived under the assumption that the unchanged class in the absolute-valued difference image follows the Half-Normal distribution, and the changed class follows the Gaussian distribution. In the spatial analysis procedure, a new conditional random fields (CRF) method based on Half-Normal distribution is proposed to model the mutual influences among image pixels. The proposed CRF method is called Half-Normal CRF (HNCRF). Experiments carried out on both synthetic datasets and four real SAR datasets demonstrate the superiority of our method. Not only few changed datasets but datasets with lots of changes are used in the experiments. The Kappa coefficients of the proposed method can reach up to ten times that of the traditional method under extreme conditions. The results prove that the proposed method outperforms the traditional methods in the case of few changed areas. Meanwhile, the proposed method can get similar results compared to traditional methods under normal conditions.

Two-dimensional prediction of the interface of geological formations: A comparative study

The location of the interface of geological formations is an important piece of information for tunneling construction. As site investigation data are usually limited, the uncertainties in locating geological interfaces for the sections between boreholes can be large and challenging to estimate. A suitable geostatistical method is thus needed for spatial prediction of the geological interfaces. In this paper, the performance of three commonly used spatial prediction methods, namely the multivariate adaptive spline regression (MARS), conditional random field (CRF) method, and thin-plate spline interpolation (TPSI) methods, are evaluated for two-dimensional cases using the boreholes data from three sites in Singapore. The prediction accuracies, patterns of the predicted surfaces, and prediction uncertainties obtained from the three methods are compared. A zonation is also proposed to improve the prediction accuracy of the MARS method. The results indicate that the MARS method can show the spatial trend of the geological interface more clearly than the other two methods. The TPSI method produces undesirable oscillations of the surface of geological interfaces and the CRF method may underestimate the extreme values of the geological interface elevations. In general, the prediction accuracy of the MARS method is similar to that of the CRF method, but higher than that of the TPSI method. For cases with very limited data in geologically complex areas, the MARS may have larger errors than the CRF method. However, the accuracy of the former can be significantly improved if a reasonable zonation is performed.

A Multiscale Joint Deep Neural Network for Glacier Contour Extraction

Rapid and accurate acquisition of glacier regional changes is of great significance to the study of glaciers. Among all satellite images, Synthetic Aperture Radar (SAR) data has a great advantage in monitoring the glaciers in harsh weather conditions. Conventionally, glacier boundaries are manually delineated on images. However, this is a time-consuming process, especially in the batch process of large-area data. In this paper, we propose a Multiscale Joint Deep Neural Network (MJ-DNN) for large-scale glaciers contour extraction using single-polarimetric SAR intensity images. Based on U-Net, the proposed method has been improved in three aspects. First, Atrous Separable Convolution is used instead of convolution with the down-sampling part. Second, we propose a multiscale joint convolution layer to obtain information at multiple scales. Third, we deepen the network with the residual connection structure for higher-level features. At the final layer, we optimize the network result with the conditional random field method. To validate our approach, we test it on three glaciers and we compare the segmentation results of four different methods in parallel. The results show that the intersection over the union of the proposed method is the most efficient.

Effect of autocorrelation function model on spatial prediction of geological interfaces

This study evaluated the performances of various autocorrelation function (ACF) models in predicting the geological interface using a well-known conditional random field method. Prediction accuracies and uncertainties were compared between a flexible Matérn model and two classical ACF models: the Gaussian model and the single exponential model. The rockhead data of Bukit Timah granite from boreholes at two sites in Singapore as well as simulated data were used for the comparisons. The results showed that the classical models produce a reasonable prediction uncertainty only when its smoothness coefficient is consistent with that of the geological data. Otherwise, the classical models may produce prediction errors much larger than that of the Matérn model. On the other hand, the prediction accuracy of the Matérn model is affected by the spacing of the boreholes. When the borehole spacing is relatively small (< 0.4 × scale of fluctuation), the Matérn model can reasonably quantify the prediction uncertainty. However, when the borehole spacing is large, the prediction by the Matérn model becomes less accurate as compared with the prediction using the classical models with the right value of smoothness coefficient due to the large estimation error of the smoothness coefficient.

Multi-class medical image segmentation using one-vs-rest graph cuts and majority voting.

Purpose: Semi-automatic image segmentation is still a valuable tool in clinical applications since it retains the expert oversights legally required. However, semi-automatic methods for simultaneous multi-class segmentation are difficult to be clinically implemented due to the complexity of underlining algorithms. We purpose an efficient one-vs-rest graph cut approach of which the complexity only grows linearly as the number of classes increases. Approach: Given an image slice, we construct multiple one-vs-rest graphs, each for a tissue class, for inference of a conditional random field (CRF). The one-vs-rest graph cut is to minimize the CRF energy derived from regional and boundary class probabilities estimated from random forests to obtain a one-vs-rest segmentation. The final segmentation is obtained by fusing from those one-vs-rest segmentations based on majority voting. We compare our method to a well-used multi-class graph cut method, alpha-beta swap, and a fully connected CRF (FCCRF) method, in brain tumor segmentation of 20 high-grade tumor cases in 2013 MICCAI dataset. Results: Our method achieved mean Dice score of 0.83 for whole tumor, compared to 0.80 by alpha-beta swap and 0.79 by FCCRF. There was a performance improvement over alpha-beta swap by a factor of five. Conclusions: Our method utilizes the probabilistic-based CRF which can be estimated from any machine learning technique. Comparing to traditional multi-class graph cut, the purposed one-vs-rest approach has complexity that grows only linearly as the number of classes increases, therefore, our method can be applicable for both online semi-automatic and offline automatic segmentation in clinical applications.

Poisson-Skellam distribution based regularization conditional random field method for photon-limited Poisson image denoising

Photon-limited Poisson image denoising is urgent demand in many application fields, but is particularly challenging because the image structures are often damaged seriously. The effective using of the image prior is very important for improving the quality of the denoised image. In this paper, we exploit the prior of inner statistical relationships among pixels in local spatial neighborhood of the Poisson noisy image and introduce a new Skellam distribution based inner interaction potential function as a distance measurement between the pixels. In the framework of conditional random field (CRF) modeling, we propose a novel Poisson-Skellam distribution based regularization CRF model for photon-limited Poisson noise removal. Using the alternating direction method of multipliers technique, we extend our method to a flexible plug-and-play scheme in which we can combine powerful Gaussian denoising method to improve the denoising performance better.

A new conditional random field based on mixture of generalized Gaussian model for synthetic aperture radar image segmentation

ABSTRACT In this paper, we propose a new algorithm using a conditional random field (CRF) model based on texture features for Synthetic Aperture Radar (SAR) image segmentation. Using the benefit of contourlet transform in describing the texture of SAR image, first, we extract contourlet coefficients from the image. Then, to take advantage of the simultaneous use of low- and high-frequency contourlet subbands, we apply the mixture of generalized Gaussian model (MoGGM) on all contourlet sub-bands to obtain more accurate texture features. Suggesting a new unary potential function based on MoGGM in our proposed CRF, we no longer require estimating the parameters in the multinominal logistic regression (MLR) model in previous CRF methods. Furthermore, we utilize the fifth-order moment to estimate the parameters of MoGGM for achieving better statistical properties of a specific region in SAR images and improving the accuracy of the texture recognition process. The experimental analysis demonstrates that segmentation results using the proposed algorithm are effectively improved compared to the previous CRF methods for SAR image segmentation.

Resume extraction with conditional random field method

Information extraction resume is a system used to carry out the automatic information extraction process on resumes of prospective employees to obtain key information including information on names, skills, educational experience, work experience, awards which are the main components in a curriculum vitae of prospective employees. Conditional Random Field (CRF) is one method that can be used to extract information. The step taken in the information extraction process is an annotation documents, tokenization, labeling on each token in the document, feature extraction, and the formation of models that will be used at the classification stage. In this study a system was built that could be used to extract CVs from users in pdf format. This study uses a CV from LinkedIn. Information extraction is done using 15 features and 11 classes. The evaluation of information extraction system for the employee’s resume has a precision value of 85,052%, a recall value of 85,052%, and an f-measure of 78,527%.

An enhanced opinion retrieval approach via implicit feature identification

Recently, there has been an enormous increase in the number of reviews of popular products. Therefore, opinion analysis has become a tedious task for customers when making decisions. As a result, opinion retrieval systems have emerged as an effective tool to analyze and represent customers’ feelings toward offered services. Conventional opinion retrieval systems retrieve and rank products according to both relevance and the overall polarity scores of the opinions. However, customer reviews are usually more detailed, including multiple features with different polarities. Consequently, feature-based opinion retrieval is necessary to extract and analyze each feature separately. Customers’ opinions are usually written with a short and unclear structure and contain many implicit linguistic features that cannot be identified by retrieval systems. As a result, the recall results are negatively affected. Few studies have focused on implicit features, as most examined explicit features. Also, implicit features extraction is a challenging task in some languages like Arabic due to difficulties with morphology. This paper proposes an enhanced retrieval approach based on feature-based opinion mining to enhance retrieval performance. In addition to explicit feature extraction, a metaheuristic optimization method with several similarity measures is utilized to identify implicit features and measure its effect on the retrieval results. The experimental results on Arabic and English datasets revealed the effectiveness of the proposed approach, whereby more features were extracted compared to the explicit feature results. Furthermore, the ranking results were improved by identifying both implicit and explicit features compared to the results obtained by the conditional random field method and association rule mining.

Adverse drug event detection using reason assignments in FDA drug labels.

Adverse drug events (ADEs) are unintended incidents that involve the taking of a medication. ADEs pose significant health and financial problems worldwide. Information about ADEs can inform health care and improve patient safety. However, much of this information is buried in narrative texts and needs to be extracted with Natural Language Processing techniques, in order to be useful to computerized methods. ADEs can be found on drug labels, contained in the different sections such as descriptions of the drug's active components or more prominently in descriptions of studied side-effects. Extracting these automatically could be useful in triaging and processing drug reports. In this paper, we present three base methods consisting of a Conditional Random Field (CRF), a bi-directional Long Short Term Memory unit with a CRF layer (biLSTM+CRF), and a pre-trained Bi-directional Encoder Representations from Transformers (BERT) model. We also present several ensembles of the CRF and biLSTM+CRF methods for extracting ADEs and their Reason from FDA drug labels. We show that all three methods perform well on our task, and that combining the models through different ensemble methods can improve results, providing increases in recall for the majority class and improving precision for all other classes. We also show the potential of framing ADE extraction from drug labels as a multi-class classification task on the Reason, or type, of ADE.

Comparative spatial predictions of the locations of soil-rock interface

The location of soil-rock interfaces (SRIs) may significantly affect the underground construction works, including the design of underground geotechnical structures. The prediction of the location of SRI using limited borehole data is a challenging task. To address this challenge, this paper presents a comparison study of four methods for spatially predicting the SRI elevation, namely the polynomial regression, spline interpolation, one-dimensional spline regression, and a Bayesian-based conditional random field. The consistencies, prediction accuracies, patterns of the predicted curves, and prediction uncertainties for various methods are evaluated. Borehole data from two sites in Singapore are used in the comparative study. The results show that the spline interpolation method produces the least consistent estimation of SRI profiles. The spline interpolation method also has lower prediction accuracies than the other three methods and cannot provide any information regarding the prediction uncertainty. The spatial trend of the geological interface cannot be captured by the polynomial regression method with a relatively high (i.e., 10) order of the polynomial when faults and folds exist. Advantages of the spline regression method over the conditional random field methods include that (i) it provides a clear and explicit spatial trend of the SRI, which well reflects the geological complexity of the sites; (ii) it avoids the cumbersome estimation of random field parametric values, which is a challenging task under the condition of limited data; and (iii) it can differentiate the zones with different prediction accuracies, which cannot be accomplished by the conditional random field method due to limited data. To sum up, the spline regression method produces a simpler and more informative curve of the SRI than the other three methods and thereby is useful as it can guide site investigations to be carried out at geologically uncertain areas to reduce risks, especially for underground construction projects

An End-to-End and Localized Post-Processing Method for Correcting High-Resolution Remote Sensing Classification Result Images

Since the result images obtained by deep semantic segmentation neural networks are usually not perfect, especially at object borders, the conditional random field (CRF) method is frequently utilized in the result post-processing stage to obtain the corrected classification result image. The CRF method has achieved many successes in the field of computer vision, but when it is applied to remote sensing images, overcorrection phenomena may occur. This paper proposes an end-to-end and localized post-processing method (ELP) to correct the result images of high-resolution remote sensing image classification methods. ELP has two advantages. (1) End-to-end evaluation: ELP can identify which locations of the result image are highly suspected of having errors without requiring samples. This characteristic allows ELP to be adapted to an end-to-end classification process. (2) Localization: Based on the suspect areas, ELP limits the CRF analysis and update area to a small range and controls the iteration termination condition. This characteristic avoids the overcorrections caused by the global processing of the CRF. In the experiments, ELP is used to correct the classification results obtained by various deep semantic segmentation neural networks. Compared with traditional methods, the proposed method more effectively corrects the classification result and improves classification accuracy.