Hybrid Bayesian Network Model Research Articles

Prognostic models for breast cancer: based on logistics regression and Hybrid Bayesian Network

BackgroundTo construct two prognostic models to predict survival in breast cancer patients; to compare the efficacy of the two models in the whole group and the advanced human epidermal growth factor receptor-2-positive (HER2+) subgroup of patients; to conclude whether the Hybrid Bayesian Network (HBN) model outperformed the logistics regression (LR) model.MethodsIn this paper, breast cancer patient data were collected from the SEER database. Data processing and analysis were performed using Rstudio 4.2.0, including data preprocessing, model construction and validation. The L_DVBN algorithm in Julia0.4.7 and bnlearn package in R was used to build and evaluate the HBN model. Data with a diagnosis time of 2018(n = 23,384) were distributed randomly as training and testing sets in the ratio of 7:3 using the leave-out method for model construction and internal validation. External validation of the model was done using the dataset of 2019(n = 8128). Finally, the late HER2 + patients(n = 395) was selected for subgroup analysis. Accuracy, calibration and net benefit of clinical decision making were evaluated for both models.ResultsThe HBN model showed that seventeen variables were associated with survival outcome, including age, tumor size, site, histologic type, radiotherapy, surgery, chemotherapy, distant metastasis, subtype, clinical stage, ER receptor, PR receptor, clinical grade, race, marital status, tumor laterality, and lymph node. The AUCs for the internal validation of the LR and HBN models were 0.831 and 0.900; The AUCs for the external validation of the LR and HBN models on the whole population were 0.786 and 0.871; the AUCs for the external validation of the two models on the subgroup population were 0.601 and 0.813.ConclusionThe accuracy, net clinical benefit, and calibration of the HBN model were better than LR model. The predictive efficacy of both models decreased and the difference was greater in advanced HER2 + patients, which means the HBN model had higher robustness and more stable predictive performance in the subgroup.

A Study on Industrial Heritage Renewal Strategy Based on Hybrid Bayesian Network

A more scientific, objective, and reasonable renewal orientation is gradually becoming a research hotspot in the field of industrial heritage conservation and renewal. This study selected five samples to carry out field investigation and face-to-face interviews. POI data were collected and analyzed, which revealed the relationship between environmental resources and the five samples based on the kernel density estimation method. Sequentially, we unprecedentedly created a complete BN-POI-AHP hybrid Bayesian network model that was used to implement simulation analysis of the industrial heritage of the Former Site Museum of Changchun Film Studio. A renewal orientation and a strategy for the community comprehensive sports ground were determined based on the results of the simulation data through the previous model. We eventually achieved a sustainable renewal strategy and innovative research method for industrial heritage, from objective data collection and simulation model creation to generation of a final reasonable plan.

Comprehensive Risk Assessment of Urban Wastewater Reuse in Water Supply Alternatives Using Hybrid Bayesian Network Model

Comprehensive Risk Assessment of Urban Wastewater Reuse in Water Supply Alternatives Using Hybrid Bayesian Network Model

Simulation and evaluation of agricultural water distribution and delivery systems with a Hybrid Bayesian network model

Regular appraisal performance of the irrigation water distribution systems from the adequacy, equity, and sustainable operation perspectives are essential due to the significant water losses that has been led by the improper adjustments of the hydraulic structures. Appropriate operational strategies could be adapted when the irrigation district authorities have a pragmatic approach to assess the hydraulic behavior of the system. This study develops an intelligent model for i) hydraulic simulation and ii) operational performance evaluation of the agricultural water distribution systems by employing a Hybrid Bayesian Networks (HBNs). The developed model was tested on a real test case located in Iran and under the different operational scenarios. In this regard, firstly, a hydrodynamic model of the test case was developed, calibrated, and validated to simulate the hydraulic condition of the irrigation canal based on a wide range of the canal inflow variations, called operational scenarios. Then, performance assessment was conducted for the scenarios based on adequacy, equity, and efficiency of water distribution perspectives. The results were employed for the training of the HBNs model. The HBNs model configuration was selected based on the irrigation district manages expectations, where the performance indicators of "adequacy," "efficiency" and "equity" of water delivery, needs to be calculated for the canal inflow variations in different operational scenarios. Validation of the developed HBNs model reveals that Mean Absolute Percentage Error (MAPE) and Coefficient of Determination (R2) of the superior model within the entire output nodes are 17.31% and 0.96, respectively. The results indicated the ability of the HBNs Model for the "Simulation/Performance Appraisal" of any operating systems in the irrigation districts. The developed model is a useful tool for i) periodic appraisal performance of the distribution system, ii) operational personals and ditch-riders, and iii) regionalization of the district to facilitate the recurrent inspections and field visiting.

An Operational Adjustment Framework for a Complex Industrial Process Based on Hybrid Bayesian Network

The operational variables used to adjust the control level of the copper cleaner flotation process have a noticeable impact on the object variables, e.g., the copper concentrate grade. Currently, due to the complexity of the flotation process, the operational variables, which are controlled by operators, are often not adjusted properly in time. Hence, this article investigates an intelligent operational adjustment framework based on a hybrid Bayesian network (BN). The offline BN model structure and the parameters are established based on process knowledge and real industrial data, respectively. After receiving the expected value of the copper concentrate grade as evidence, an operational adjustment can be obtained online by BN reasoning. To ensure its credibility, the copper concentrate grade after operational adjustment is further predicted. According to the predicted value, the operators can determine whether to implement the operational adjustment or not. Finally, the experimental results show the effectiveness and practical significance of the proposed method. Note to Practitioners— This article was motivated by the problem of operational adjustment in the copper cleaner flotation process. Because of the current manual operational adjustment, controlling the concentrate grade within its acceptable range is difficult. This article suggests a hybrid Bayesian network (BN) approach to address this problem. The hybrid BN model is built offline and used to make inferences online. Compared with a conventional discrete BN, this approach can provide operators with specific adjustment values to qualify the concentrate grade. For the evaluation, data experiments are used. This evaluation shows the efficiency and superiority of the proposed approach in terms of automation, intelligence, and decision-making.

Development of a hybrid Bayesian network model for predicting acute fish toxicity using multiple lines of evidence

A hybrid Bayesian network (BN) was developed for predicting the acute toxicity of chemicals to fish, using data from fish embryo toxicity (FET) testing in combination with other information. This model can support the use of FET data in a Weight-of-Evidence (WOE) approach for replacing the use of ju-venile fish. The BN predicted correct toxicity intervals for 69%–80% of the tested substances. The model was most sensitive to components quantified by toxicity data, and least sensitive to compo-nents quantified by expert knowledge. The model is publicly available through a web interface. Fur-ther development of this model should include additional lines of evidence, refinement of the discre-tisation, and training with a larger dataset for weighting of the lines of evidence. A refined version of this model can be a useful tool for predicting acute fish toxicity, and a contribution to more quantitative WOE approaches for ecotoxicology and environmental assessment more generally.

Real-time risk assessment of a fracturing manifold system used for shale-gas well hydraulic fracturing activity based on a hybrid Bayesian network

A fracturing manifold system used in the shale-gas hydraulic fracturing process involves significant risks, which result from its unique and harsh working conditions, such as high pressures of up to 105 MPa and a large displacement, along with the continuous erosion and corrosion of high-speed solid particles. Considering the high- or low-frequency demand modes of the components and the effect of any deviation in the state indicators on the real-time risk of the fracturing manifold system, we propose a real-time risk assessment method based on a hybrid Bayesian network (HBN) to provide decision support for supervisors that will prevent accidents. The proposed approach can be utilised to calculate the probability of each potential consequence in real time. The built HBN model was quantified by using the historical failure-related data of various components, specific monitoring data of multiple state indicators and expert judgment. An extensive case study focused on the real-time risk of a real-world fracturing manifold system and demonstrated the practical application of the presented methodology. We show by application that the proposed model can improve the situational awareness among operators, which is an effective method to control and reduce risk.

A hybrid Bayesian network model for predicting delays in train operations

• Bayesian network is introduced to tackle delays in train operations. • Train operation data is used to learn the BN structures. • The heuristic-based structure is elaborated as a prediction model. • The hybrid BN model is tested using different performance measures. We present a Bayesian network-(BN) based train delay prediction model to tackle the complexity and dependency nature of train operations. Three different BN schemes, namely, heuristic hill-climbing, primitive linear and hybrid structure, are investigated using real-world train operation data from a high-speed railway line. We first use historical data to rationalize the dependency graph of the developed structures. Each BN structure is then trained with the gold standard k -fold cross validation approach to avoid over-fitting and evaluate its performance against the others. Overall, the validation results indicate that a BN-based model can be an efficient tool for capturing superposition and interaction effects of train delays. However, a well-designed hybrid BN structure, developed based on domain knowledge and judgments of expertise and local authorities, can outperform the other models. We present a performance comparison of the predictions obtained from the hybrid BN structure against the real-world benchmark data. The results show that the proposed model on overage can achieve over 80% accuracy in predictions within a 60-min horizon, yielding low prediction errors regarding mean absolute error (MAE), mean error (ME) and root mean square error (RMSE) measures.

A Probabilistic Nonlinear Model for Forecasting Daily Water Level in Reservoir

Accurate prediction and monitoring of water level in reservoirs is an important task for the planning, designing, and construction of river-shore structures, and in taking decisions regarding irrigation management and domestic water supply. In this work, a novel probabilistic nonlinear approach based on a hybrid Bayesian network model with exponential residual correction has been proposed for prediction of reservoir water level on daily basis. The proposed approach has been implemented for forecasting daily water levels of Mayurakshi reservoir (Jharkhand, India), using a historic data set of 22 years. A comparative study has also been carried out with linear model (ARIMA) and nonlinear approaches (ANN, standard Bayesian network (BN)) in terms of various performance measures. The proposed approach is comparable with the observed values on every aspect of prediction, and can be applied in case of scarce data, particularly when forcing parameters such as precipitation and other meteorological data are not available.

A Bayesian Network Model of the Particle Swarm Optimization for Software Effort Estimation

Rapid growth of software industry leads to need of new technologies. Software effort estimation is one of the areas that need more attention. Exact estimation is always a challenging task. Effort Estimation techniques are broadly classified into algorithmic and non-algorithmic techniques. An algorithmic model provides a mathematical equation for estimation which is based upon the analysis of data gathered from previously developed projects and Non-algorithmic techniques are based on new approaches, such as Soft Computing Techniques. Effective handling of cost is a basic need for any Software Organization. This paper presents the new hybrid Bayesian Network model of PSO for effort estimation. we have developed a tool in MATLAB and at last proved that Bayesian Network with PSO gives more accurate results than other existing techniques. For sake of ease, we use NASA 93 datasets to verify the model and also compare the proposed model with COCOMO and Bayesian Regulation Neural Network Model and it is found that the developed model provides better estimation. General Terms Particle Swarm Optimization, Neural network, Bayesian Network.

A Hybrid Bayesian Network Model for Predicting Breast Cancer Prognosis

Objective: Breast cancer is one of the most common cancers affecting women. Both physicians and patients have concerned about breast cancer survivability. Many researchers have studied the breast cancer survivability applying artificial nerural network model (ANN). Usually ANN model outperformed in classification of breast cancer survivability than other models such as logistic regression, Bayesian network (BN), or decision tree models. However, physicians in the fields hesitate to use ANN model, because ANN is a black-box model, and hard to explain the classification result to patients. In this study, we proposed a hybrid model with a degree of the accuracy and interpretation by combining the ANN for accuracy and BN for interpretation. Methods: We developed an artificial neural network, a Bayesian network, and a hybrid Bayesian network model to predict breast cancer prognosis. The hybrid model combined the artificial neural network and the Bayesian network to obtain a good estimation of prognosis as well as a good explanation of the results. The National Cancer Institute's SEER program public-use data (1973-2003) were used to construct and evaluate the proposed models. Nine variables, which are clinically acceptable, were selected for input to the proposed models' nodes. A confidence value of the neural network served as an additional input node to the hybrid Bayesian network model. Ten iterations of random subsampling were performed to evaluate performance of the models. Results: The hybrid BN model achieved the highest area under the curve value of 0.935, whereas the corresponding values of the neural network and Bayesian network were 0.930 and 0.813, respectively. The neural network model achieved the highest prediction accuracy of 88.8% with a sensitivity of 93.7% and a specificity of 85.4%. The hybrid Bayesian network model achieved a prediction accuracy of 87.2% with a sensitivity of 93.3% and a specificity of 83.1%. The results of the hybrid Bayesian network model were very similar to the neural network model. Conclusion: In the experiments, the hybrid model and the ANN model outperformed the Bayesian network model. The proposed hybrid BN model for breast cancer prognosis predictin may be useful for clinicians in the medical fields, as the model provides both high degree of performance inherited from ANN and good explanation power from BN. (Journal of Korean Society of Medical Informatics 15-1, 49-57, 2009)