Graph-based Method Research Articles

A Structure-aware Hierarchical Graph-based Multiple Instance Learning Framework for pT Staging in Histopathological Image.

Pathological primary tumor (pT) stage focuses on the infiltration degree of the primary tumor to surrounding tissues, which relates to the prognosis and treatment choices. The pT staging relies on the field-of-views from multiple magnifications in the gigapixel images, which makes pixel-level annotation difficult. Therefore, this task is usually formulated as a weakly supervised whole slide image (WSI) classification task with the slide-level label. Existing weakly-supervised classification methods mainly follow the multiple instance learning paradigm, which takes the patches from single magnification as the instances and extracts their morphological features independently. However, they cannot progressively represent the contextual information from multiple magnifications, which is critical for pT staging. Therefore, we propose a structure-aware hierarchical graph-based multi-instance learning framework (SGMF) inspired by the diagnostic process of pathologists. Specifically, a novel graph-based instance organization method is proposed, namely structure-aware hierarchical graph (SAHG), to represent the WSI. Based on that, we design a novel hierarchical attention-based graph representation (HAGR) network to capture the critical patterns for pT staging by learning cross-scale spatial features. Finally, the top nodes of SAHG are aggregated by a global attention layer for bag-level representation. Extensive studies on three large-scale multi-center pT staging datasets with two different cancer types demonstrate the effectiveness of SGMF, which outperforms state-of-the-art up to 5.6% in the F1 score.

Knowledge graph-based recommendation method for cold chain logistics

Knowledge graph-based recommendation method for cold chain logistics

Study on nonlinear dynamics characteristics of dual speed dual clutch transmission system based on bond graph

The variable speed transmission system employs compound planetary gear trains and clutches. Compared to traditional aviation transmission systems, the added control parts in the variable transmission system result in a stronger coupling between components and more complex dynamic response characteristics. Modeling using the central mass method is complex and abstract. This study proposes a bond graph-based method for modeling and analyzing the dynamic behavior of the variable transmission system. The transmission system is divided into multiple component-level real-time dynamic models, and obtain a nonlinear multi-energy domain shift dynamic models based on energy flow paths coupling. The mathematical model is numerically solved using the Runge-Kutta method. Simulation results show that there is a delay in the speed response during the shifting process. Further analysis reveals that this “give and take” phenomenon is related to the mechanical structure and passive working principle of the one-way clutch. Subsequently, the time-domain response of the output speed under the hydraulic loading characteristic curves of linear, exponential, S-shaped, and composite functions, as well as the clutch torque, were studied. The impact of different hydraulic loading durations based on the exponential curve on component speed and torque was also analyzed. The simulation results provide a theoretical basis for designing low-shift impact, high-reliability, and high-performance variable transmission systems.

Robust multi-step wind speed forecasting based on a graph-based data reconstruction deep learning method

Against global warming, wind energy has increasingly become a stable form of power supply. Accurate prediction of wind speed is crucial for turbine control and wind farm dispatch, contributing to stable and continued wind energy utilization. However, it is very difficult to accomplish satisfactory wind speed forecasting, especially multi-step forecasting, due to the stochastic, random and volatile characteristic of wind speed series with complex fluctuations. This study is elaborated to propose a novel method based on “data graph” reconstruction and long short term memory (LSTM) network, to achieve accurate and robust multi-step wind speed forecasting. To obtain implicit correlations between wind speed series, the time series data are reconstructed into matrices like “data graphs”. To achieve better computing efficiency and forecasting accuracy, the convolutional neural network (CNN) is adopted to extract the features in the “data graphs”. Then the extracted data graph features are imported into the bidirectional-LSTM (bi-LSTM) network module, which takes the input in forward and backward directions, respectively, to extract more temporal information. For adequate performance assessment, experiments are carried out on data sets of an actual wind farm located in a mountainous region in China. The results show that the proposed method outperforms classic meta-model based and hybrid-model based methods in accuracy, stability and computing efficiency. The results reveal that the data graph reconstruction combined with the CNN is able to extract the hidden features of the wind speed time series data. The proposed CNN-bi-LSTM method effectively improve the multi-step wind speed prediction accuracy.

Abdominal electromyograms mining: breathing patterns of asleep adults

The article’s subject matter is the processing of abdominal EMG recordings and finding breathing patterns. The goal is to automatically classify respiratory patterns into two classes, or clusters, by two breathing patterns, regular and irregular, using machine learning (ML) methods. The object of the study was to obtain a dataset of 40 randomly picked abdominal EMG recordings (sampling rate equal to 200 Hz) borrowed from the complete dataset published by the Computational Clinical Neurophysiology Laboratory and the Clinical Data Animation Laboratory of Massachusetts General Hospital. The tasks to be solved are as follows: finding ETS (errors-trend-seasonality) model for the EMG series using the exponential smoothing method; obtaining denoised and detrended signals; obtaining the Hurst exponents for EMGs using the power-law decaying of correlograms for the denoised and detrended signals; describing the variabilities, SNR, the outlier fractions, and Hurst exponents by robust statistics, performing correlation analysis, and Principal Components Analysis (PCA); analyzing the structure of the distant matrix by a graph-based technique; obtaining the periodograms in the frequency domain using the known Wiener-Khinchin theorem; and finding the best models and methods of classification and clusterization and evaluating them within modern Machine Learning methods. The methods used are exponential smoothing, the Wiener-Khinchin theorem, the graph theory method, principal component analysis, programing within MAPLE 2020, and data processing by Weka. The authors obtained the following results: 1) wide data variability has been rated with the median absolute deviations, which is the most robust statistic in this case; 2) most of the signals (38 of 40) showed frequent outliers: from a few percent up to 24.6 % of emissions; 3) these four variables: outliers' percentage, variability, SNR, and persistency factors – form the attributes of input vectors of the subjects for further Machine Learning with Weka software; 4) Manhattan distances matrix among subjects' vectors in 4D attributes space allows imaging the data set as a weighted graph, the vertices of which are subjects; 5) the weights of the graph's edges reflect distances between any pair of them. "Closeness centralities" of vertices allowed us to cluster the data set on two clusters with 11 and 29 subjects, and Weka clustering algorithms confirmed this result. 6) The learning curve shows that a sufficiently small data set (from 25 subjects) might be suitable for classification purposes. Conclusions. The scientific novelty of the results obtained is as follows: 1) the Error-Trend-Seasonality model was the same for all data sets. Abdominal EMG of sleeping patients had additive errors and undamped trends without any seasonality; 2) the correlograms' decaying according to power law had been set, and Hurst exponents were in the range (of 0.776–0.887). This testifies to "long memory" (high persistence) of abdominal EMGs; 3) the modified Z-scores and robust statistics with the highest breakdown values were used for the EMG parameters because of many outliers; 4) breathing patterns were set using the periodograms in the frequency domain using the Wiener-Khinchin theorem; 5) the new graph-based method was successfully exploited to cluster the dataset. Parallel clustering with Weka algorithms confirmed the graph-based clustering results.

A systematic knowledge graph-based smart management method for operations: A case study of standardized management

Standardized routine operation management (SROM) has been widely accepted and applied by kinds of enterprises and played a key supporting role. With full use of the emerging knowledge-based smart management technology, SROM will further increase comprehensive efficiency and save human resources greatly at the same time, especially for small and medium enterprises (SMEs). Hence, we propose a systematic knowledge-based smart management method to transfer SROM activities from human operations to automatic response by means of knowledge explicitation, organization, sharing and reusing, which can be further achieved by employing knowledge graph. We took a typical SROM instance, ISO 9000 implementation management, as an example to validate the transformation from human activities to knowledge graph-based automatic operation. We firstly analyzed characteristics of domain knowledge and constructed an ontology model according to the knowledge stability. Secondly, a hybrid knowledge graph construction and dynamic updating framework together with related algorithms were designed by deliberately integrating semantic similarity calculation and natural language processing. Thirdly, we developed a question-answering mechanism and reasoning system based on the ISO 9000 implementation knowledge graph to support automatic decision and feedback for ISO 9000 routine operation management including knowledge learning and processes auditing. Finally, the practicability and effectiveness of SROM knowledge graph has been validated in a SME in China, realizing the application of question-answering, job responsibility recommendation, conflict detection, semantic detection, multidimensional statistical analysis. The proposed method can also be generalized to support auxiliary optimization decision, vertical risk control, operation mode analysis, optimization model improvement experience and so on.

Fault diagnosis method for hydro-power plants with Bi-LSTM knowledge graph aided by attention scheme

In hydro-power systems, the fault of equipment is an important potential threat for the safe production of electricity. Therefore, the automation and intelligence of fault diagnosis becomes the popular issue in the research on hydro-power system. In this paper, a knowledge graph-based method is put forth to diagnose faults occurred in hydro-power systems, since the knowledge graph can store structured and unstructured data for better fault diagnosis and intelligently search the reasons of the faults. First, we model the knowledge graph for hydro-power plants, where the rational path for the fault reason is formulated. Then, the bi-directional long short-term memory (Bi-LSTM) with conditional random field (CRF) is used to extract the entities and relations to the given documents, which record the phenomenon and reasons for the occurred faults. Moreover, the attention scheme is employed in the Bi-LSTM to weigh the closer relationships to improve the diagnosis accuracy. An automatic diagnosis algorithm is developed to improve the diagnosing efficiency by constructing rational paths, with which directive and in-directive factors for occurring faults can be traced. Simulation results reveal that the intelligent search method with a knowledge graph can effectively find the reason, locate the position, and provide useful suggestions for the occurred faults.

Assessing geometrical uncertainties in geological interface models using Markov chain Monte Carlo sampling via abstract graph

Uncertainty assessment is a common requirement in 3D modeling applications. The Markov chain Monte-Carlo (MCMC) method is a practical way to generate multiple realizations of 3D models to evaluate model uncertainty. However, when probing high-dimensional target distributions related to the detailed geometry of 3D geological interfaces, the generating 3D models by existing MCMC methods always suffer from the presence of artifacts in large geometrical deformation that modeling deformed geometries in stochastic perturbation, which limits their effectiveness and efficiency. This paper presents an abstract graph-based MCMC method for the uncertainty assessment of 3D geological interfaces. An abstract graph is used to represent the reasonable shape of the geological interface. In proposing new candidate 3D models via the MCMC iterations, the 3D models are perturbed such that the energy function measuring the distortion of the abstract graph is minimized in a least-squares sense. This can preserve the plausible geometry of the 3D models against high perturbations and provide artifact-free 3D models as candidates for MCMC sampling. The resulting MCMC iterations enable the drawing of probable realizations of 3D models while maintaining a high acceptance rate in probing the complicated target distributions of 3D geological interface models. The proposed method was applied to the assessment of uncertainties in three regional faults in the Jiaodong Peninsula, eastern China. The uncertainty assessment was carried out based on a complex probability distribution with consideration of geological-geophysical observations and prior knowledge of fault geometry. The results show that the proposed MCMC method could reasonably locate the uncertainties in free-form fault models defined by thousands of constraints. And the abstract graph-based MCMC method can sample highly-perturbed and geometrically plausible 3D models in high-resolution, which is crucial to the effectiveness and efficiency of the geometrical uncertainty assessment.

Extracting Broken-Rotor-Bar Fault Signature of Varying-Speed Induction Motors

An effective way to detect broken-bar faults of squirrel-cage induction motors is to extract the characteristic frequency component in the stator current as a fault signature, or so-called motor current signature analysis (MCSA). However, for inverter-fed motor drive systems, the motor is typically operating under varying-speed, varying-load, and noisy environments, which makes the fault signature extraction a very challenging problem. In this paper, we propose a sparsity-driven and graph-based method to extract the fault signature effectively, where the fault signature is modeled as a sparse component in the frequency domain for each short-time window measurement while gradually changing from window to window in the time-domain. Compared to the conventional short-time Fourier transform-based method, our method is more robust to noise and varying speed operations. Experiments are carried out to demonstrate the effectiveness of the proposed method.

Facilitating Knowledge Transfer during Code Compliance Checking Using Conceptual Graphs

The rules in building codes are composed of language, logic structure, and the embedded knowledge that requires further articulation from rule experts. This greatly increases the complexity of rule interpretation. The conceptual graph (CG) is used as a graph-based representation method to capture such knowledge and transfer it to other actors in the compliance checking process. However, due to its subjectivity, it is difficult to ensure the quality of the CG to achieve minimal knowledge loss. This paper proposes a method using reusable CG patterns to reduce the knowledge loss caused by the opacity of concepts in rules, hence better facilitating knowledge transfer using CGs. The rules and their corresponding CGs are analyzed and compared. An assessment framework is used to assess CGs for correctness, completeness, complexity, and comprehensibility. A group of reusable CG patterns is then generated based on the analysis results of concepts in existing CGs. In a case study, rules selected from two types of building codes are used to validate the proposed methods. The results successfully demonstrate the effectiveness of the CG in facilitating knowledge transfer. Moreover, the comparison of quality between original CGs and improved CGs shows that using reusable CG patterns can effectively ease the creation of new CGs and improve its consistency.

Finding Taint-Style Vulnerabilities in Lua Application of IoT Firmware with Progressive Static Analysis

With the rapid growth of IoT devices, ensuring the security of embedded firmware has become a critical concern. Despite advances in existing vulnerability discovery methods, previous research has been limited to vulnerabilities occurring in binary programs. Although an increasing number of vendors are utilizing Lua scripting language in firmware development, no automated method is currently available to discover vulnerabilities in Lua-based programs. To fill this gap, in this paper, we propose FLuaScan, a novel progressive static analysis approach specifically designed to detect taint-style vulnerabilities in Lua applications in IoT firmware. FLuaScan first heuristically locates the code that handles user input, then divides the code into different segments to conduct a progressive taint analysis. Finally, a graph-based search method is applied to identify vulnerable code that satisfies the conditions of taint propagation. To comprehensively compare FLuaScan with state-of-the-art tool Tscancode, we conducted various experiments on a dataset consisting of 13 real-world firmware samples from different vendors. The results demonstrate the superior performance of FLuaScan in terms of accuracy (increased TP rate from 0% to 42.50%), effectiveness (discovered 21 vulnerabilities, of which 7 are unknown), and practicality (acceptable time overhead and visual output to assist in manual analysis).

CAM-NAS: An Efficient and Interpretable Neural Architecture Search Model Based on Class Activation Mapping

Artificial intelligence (AI) has made rapid progress in recent years, but as the complexity of AI models and the need to deploy them on multiple platforms gradually increases, the design of network model structures for specific platforms becomes more difficult. A neural network architecture search (NAS) serves as a solution to help experts discover new network structures that are suitable for different tasks and platforms. However, traditional NAS algorithms often consume time and many computational resources, especially when dealing with complex tasks and large-scale models, and the search process can become exceptionally time-consuming and difficult to interpret. In this paper, we propose a class activation graph-based neural structure search method (CAM-NAS) to address these problems. Compared with traditional NAS algorithms, CAM-NAS does not require full training of submodels, which greatly improves the search efficiency. Meanwhile, CAM-NAS uses the class activation graph technique, which makes the searched models have better interpretability. In our experiments, we tested CAM-NAS on an NVIDIA RTX 3090 graphics card and showed that it can evaluate a submodel in only 0.08 seconds, which is much faster than traditional NAS methods. In this study, we experimentally evaluated CAM-NAS using the CIFAR-10 and CIFAR-100 datasets as benchmarks. The experimental results show that CAM-NAS achieves very good results. This not only proves the efficiency of CAM-NAS, but also demonstrates its powerful performance in image classification tasks.

Innovative method of radiotracer data analysis to identify leaky heat exchangers

The radiotracers are most widely used in the petrochemical industries to identify leaky heat exchangers. Gamma radiation transmitted by radiotracer is detected at strategic locations and recorded as data. This data is analyzed using a standard graph-based method. However, understanding the leakage problems from graphs is not always straightforward. Its success depends on the knowledge, experience, and skill of the interpreter who identifies the leakage. The unavailability of a better alternative has motivated the development of a new data analysis method. The paper describes a new innovative method of data analysis developed to identify leaky heat exchangers and quantify the leak rate. The method focuses on the analysis of time-domain data by transforming it into the frequency domain. The main objective of this alternate method is to support the determination of leaky heat exchangers from radiotracer data. It also aims to analyse data quantitatively and improve the overall performance of the investigation. A step-by-step approach to data analysis is developed by identifying specific requirements. The method is tested and evaluated successfully. It is observed that the new way of radiotracer data analysis gives better results without drawing the graphs. Moreover, it supports quantitative analysis of data rather than qualitative.

General provisions of the substantiation of functional safety of intelligent systems in railway transportation

Aim. The paper aims to solve the problem of objective and confident functional safety (FS) evaluation of intelligent control systems (ICS). As regards ICS, the conventional methods, due to their particular features, do not allow for a sufficiently confident estimation of the actual state of FS. The above features include primarily the nondistinct architecture of ICS and the changing connections between the system elements.Methods. Substantiating ICS FS requires using the complete arsenal of known methods and means recommended in GOST 33432-2015 [1], including managerial measures defined by the requirements for the safety policy, program and case. The authors have analysed the capability to prove ICS FS using experimental, expert, analytical, technological, and simulation-based methods. The limitations of some methods as regards ICS FS substantiation have been established.Results. The authors suggest a heuristic graph-based semi-Markov (Markov) method of proving system FS. For the purpose of substantiating ICS FS, it is recommended using the heuristic graph-based method combined with the technological method defined in GOST R IEC 61508 [2–4]. They don’t only allow confidently evaluating the FS of intelligent systems, but developing recommendations for achieving acceptable safety levels of such systems.

Graph-based automatic detection and classification of lesion changes in pairs of CT studies for oncology follow-up.

Radiological follow-up of oncology patients requires the quantitative analysis of lesion changes in longitudinal imaging studies, which is time-consuming, requires expertise, and is subject to variability. This paper presents a comprehensive graph-based method for the automatic detection and classification of lesion changes in current and prior CT scans. The inputs are the current and prior CT scans and their organ and lesion segmentations. Classification of lesion changes is formalized as bipartite graph matching where lesion pairings are computed by adaptive overlap-based lesion matching. Six types of lesion changes are computed by connected components analysis. The method was evaluated on 208 pairs of lung and liver CT scans from 57 patients with 4600 lesions, 1713 lesion matchings and 2887 lesion changes. Ground-truth lesion segmentations, lesion matchings and lesion changes were created by an expert radiologist. Our method yields a lesion matching rate accuracy of 99.7% (394/395) and 95.0% (1252/1318) for the lung and liver datasets. Precision and recall are > 0.99 and 0.94 and 0.95 (respectively) for the detection of lesion changes. The analysis of lesion changes helped the radiologist detect 48 missed lesions and 8 spurious lesions in the input ground-truth lesion datasets. The classification of lesion classification provides the clinician with a readily accessible and intuitive identification and classification of the lesion changes and their patterns in support of clinical decision making. Comprehensive automatic computer-aided lesion matching and analysis of lesion changes may improve quantitative follow-up and evaluation of disease status, assessment of treatment efficacy and response to therapy.

A graph-based interpretability method for deep neural networks

With the development of artificial intelligence, the most representative deep learning has been applied to various fields, which is greatly influencing human society. However, deep neural networks (DNNs) are still a black-box model, and the process how they make decisions internally is still difficult to understand and control. At the same time, DNNs take up more hardware resources, resulting in high energy consumption. Therefore, it is significant to study the characteristics of deep AI models and deeply understand the interactions between parameters within AI models so as to improve the interpretability of DNNs, optimize their structure and increase their computational efficiency. In this paper, we propose a graph-based interpretability method for deep neural networks (GIMDNN). The running parameters of DNNs are modeled as a graph by using a kernel function or the Graph Transformer Networks (GTN), where the nodes of the graph are obtained by dimensional mapping of the parameters of the DNNs, and the edges are calculated by the Gaussian kernel function. The generated graphs are classified by a graph convolutional network (GCN). The association relationship between the adjacent layers and the running mechanism of DNNs are analyzed, and the importance of the parameters of each layer in the DNNs for the final classification result can be obtained. Convolutional neural networks (CNNs) are one of the most representative models in DNNs. The proposed method is experimentally evaluated on the CNNs. The experimental results show that the proposed method can interpret the associations among the weight parameters as well as the correlation between two adjacent layers. Therefore, the DNNs for special tasks, such as portable applications, edge computing, and so on, can be customized, the number of parameters can be reduced. It is valuable to interpret the operation and principle of CNNs.

Adversarial Robustness in Graph-Based Neural Architecture Search for Edge AI Transportation Systems

Edge AI technologies have been used for many Intelligent Transportation Systems, such as road traffic monitor systems. Neural Architecture Search (NAS) is a typcial way to search high-performance models for edge devices with limited computing resources. However, NAS is also vulnerable to adversarial attacks. In this paper, A One-Shot NAS is employed to realize derivative models with different scales. In order to study the relation between adversarial robustness and model scales, a graph-based method is designed to select best sub models generated from One-Shot NAS. Besides, an evaluation method is proposed to assess robustness of deep learning models under various scales of models. Experimental results shows an interesting phenomenon about the correlations between network sizes and model robustness, reducing model parameters will increase model robustness under maximum adversarial attacks, while, increasing model paremters will increase model robustness under minimum adversarial attacks. The phenomenon is analyzed, that is able to help understand the adversarial robustness of models with different scales for edge AI transportation systems.

Kmer-Node2Vec: a Fast and Efficient Method for Kmer Embedding from the Kmer Co-occurrence Graph, with Applications to DNA Sequences.

Learning low-dimensional continuous vector representation for short k-mers divided from long DNA sequences is key to DNA sequence modeling that can be utilized in many bioinformatics investigations, such as DNA sequence retrieval and classification. DNA2Vec is the most widely used method for DNA sequence embedding. However, it poorly scales to large data sets due to its extremely long training time in kmer embedding. In this paper, we propose a novel efficient graph-based kmer embedding method, named Kmer-Node2Vec, to tackle this concern. Our method converts the large DNA corpus into one kmer co-occurrence graph, and extracts kmer relation on the graph by random walks to learn fast and high-quality kmer embedding. Extensive experiments show that our method is faster than DNA2Vec by 29 times for training on a 4GB data set, and on par with DNA2Vec in terms of task-specific accuracy of sequence retrieval and classification.

A unified semi-supervised model with joint estimation of graph, soft labels and latent subspace

Since manually labeling images is expensive and labor intensive, in practice we often do not have enough labeled images to train an effective classifier for the new image classification tasks. The graph-based SSL methods have received more attention in practice due to their convexity, scalability and efficiency. In this paper, we propose a novel graph-based semi-supervised learning method that takes full advantage of a small set of labeled graphs and a large set of unlabeled graph data. First, we explain the concept of graph-based semi-supervised learning. The core idea of these models is to jointly estimate a low-rank graph with soft labels and a latent subspace. The proposed scheme leverages the synergy between the graph structure and the data representation in terms of soft labels and latent features. This improves the monitoring information and leads to better discriminative linear transformation. Several experiments were conducted on five image datasets using state-of-the-art methods. These experiments show the effectiveness of the proposed semi-supervised method.

A Graph-Based Model Reduction Method for Digital Twins

Digital twin technology is the talking point of academia and industry. When defining a digital twin, new modeling paradigms and computational methods are needed. Developments in the Internet of Things and advanced simulation and modeling techniques have provided new strategies for building complex digital twins. The digital twin is a virtual entity representation of the physical entity, such as a product or a process. This virtual entity is a collection of computationally complex knowledge models that embeds all the information of the physical world. To that end, this article proposes a graph-based representation of the virtual entity. This graph-based representation provides a method to visualize the parameter and their interactions across different modeling domains. However, the virtual entity graph becomes inherently complex with multiple parameters for a complex multidimensional physical system. This research contributes to the body of knowledge with a novel graph-based model reduction method that simplifies the virtual entity analysis. The graph-based model reduction method uses graph structure preserving algorithms and Dempster–Shaffer Theory to provide the importance of the parameters in the virtual entity. The graph-based model reduction method is validated by benchmarking it against the random forest regressor method. The method is tested on a turbo compressor case study. In the future, a method such as graph-based model reduction needs to be integrated with digital twin frameworks to provide digital services by the twin efficiently.