A Feature Weighted Mixed Naive Bayes Model for Monitoring Anomalies in the Fan System of a Thermal Power Plant

With the increasing intelligence and integration,a great number of two-valued variables(generally stored in the form of 0 or 1)often exist in large-scale industrial processes.However,these variables cannot be effectively handled by traditional monitoring methods such as linear discriminant analysis(LDA),principal component analysis(PCA)and partial least square(PLS)analysis.Recently,a mixed hidden naive Bayesian model(MHNBM)is developed for the first time to utilize both two-valued and continuous variables for abnormality monitoring.Although the MHNBM is effective,it still has some shortcomings that need to be improved.For the MHNBM,the variables with greater correlation to other variables have greater weights,which can not guarantee greater weights are assigned to the more discriminating variables.In addition,the conditional P(x j|x j′,y=k)probability must be computed based on historical data.When the training data is scarce,the conditional probability between continuous variables tends to be uniformly distributed,which affects the performance of MHNBM.Here a novel feature weighted mixed naive Bayes model(FWMNBM)is developed to overcome the above shortcomings.For the FWMNBM,the variables that are more correlated to the class have greater weights,which makes the more discriminating variables contribute more to the model.At the same time,FWMNBM does not have to calculate the conditional probability between variables,thus it is less restricted by the number of training data samples.Compared with the MHNBM,the FWMNBM has better performance,and its effectiveness is validated through numerical cases of a simulation example and a practical case of the Zhoushan thermal power plant(ZTPP),China.

A Naive Bayes model on lung adenocarcinoma projection based on tumor microenvironment and weighted gene coexpression network analysis

Based on the lung adenocarcinoma(LUAD)gene expression data from the cancer genome atlas(TCGA)database,the Stromal score,Immune score and Estimate score in tumor microenvironment(TME)were computed by the Estimation of Stromal and Immune cells in Malignant Tumor tissues using Expression data(ESTIMATE)algorithm.And gene modules significantly related to the three scores were identified by weighted gene coexpression network analysis(WGCNA).Based on the correlation coefficients and P values,899 key genes affecting tumor microenvironment were obtained by selecting the two most correlated modules.It was suggested through Gene Ontology(GO)and Kyoto Encyclopedia of Genes and Genomes(KEGG)enrichment analysis that these key genes were significantly involved in immune-related or cancer-related terms.Through univariate cox regression and elastic network analysis,genes associated with prognosis of the LUAD patients were screened out and their prognostic values were further verified by the survival analysis and the University of ALabama at Birmingham CANcer(UALCAN)database.The results indicated that eight genes were significantly related to the overall survival of LUAD.Among them,six genes were found differentially expressed between tumor and control samples.And immune infiltration analysis further verified that all the six genes were significantly related to tumor purity and immune cells.Therefore,these genes were used eventually for constructing a Naive Bayes projection model of LUAD.The model was verified by the receiver operating characteristic(ROC)curve where the area under curve(AUC)reached 92.03%,which suggested that the model could discriminate the tumor samples from the normal accurately.Our study provided an effective model for LUAD projection which improved the clinical diagnosis and cure of LUAD.The result also confirmed that the six genes in the model construction could be the potential prognostic biomarkers of LUAD.

具有椭球等高分布误差的半参数回归模型参数的Bayes估计

本文给出了具有椭球等高分布误差的半参数回归模型中参数的Ｂａｙｅｓ估计．

A combined numerical tidal model for the Hangzhou Bay and Qiantang River

-In order to avoid prescribing open boundary condition on the upstream side of the Hangzhou Bay, in numerical simulation of the tides and residual currents of the Bay, a 1-D model for the Qiantang River is connected to the 2-D model for the Hangzhou Bay. The harmonic constants of diurnal constituent [ (K1+O1)/2],semidiurnal constituent (M2) and shallow water constituent (M4) are obtained. The results produced by the combined model are in better agreement with the observed ones than those produced solely by the original 2-D model. The combined model gives much more reliable results for tide-induced residual water level and current.

Implementing a multispecies size-spectrum model in a datapoor ecosystem

Multispecies ecological models have been used for predicting the effects of fishing activity and evaluating the performance of management strategies. Size-spectrum models are one type of physiologically-structured ecological model that provide a feasible approach to describing fish communities in terms of individual dietary variation and ontogenetic niche shift. Despite the potential of ecological models in improving our understanding of ecosystems, their application is usually limited for data-poor fisheries. As a first step in implementing ecosystem-based fisheries management（EBFM）, this study built a size-spectrum model for the fish community in the Haizhou Bay, China. We describe data collection procedures and model parameterization to facilitate the implementation of such size-spectrum models for future studies of data-poor ecosystems. The effects of fishing on the ecosystem were exemplified with a range of fishing effort and were monitored with a set of ecological indicators. Total community biomass, biodiversity index, W-statistic, LFI（Large fish index）, Mean W（mean body weight） and Slope（slope of community size spectra） showed a strong non-linear pattern in response to fishing pressure, and largest fishing effort did not generate the most drastic responses in certain scenarios. We emphasize the value and feasibility of developing size-spectrum models to capture ecological dynamics and suggest limitations as well as potential for model improvement. This study aims to promote a wide use of this type of model in support of EBFM.

The Application of Ecosystem Dynamic Model in Xiamen Bay

This paper is to establish a nitrogen and phosphorus nutrients cycle-based numerical model of ecological dynamics for Xiamen Bay on the basis of the existing three-dimensional barocline hydrodynamic model. The calculation results show that the estuarine district of Jiulongjiang estuary has the highest inorganic nitrogen concentration followed by the West Harbor, which demonstrates that Jiulongjiang River is the main input source of inorganic nitrogen in Xiamen Bay. The West Harbor has relatively high concentration of nutrients caused by the huge land pollution emission and its own poor water exchange capacity; while the distribution rules of phytoplankton biomass correspond with those of phosphates, demonstrating Xiamen Bay＇s phytoplankton controlled by phosphorus; the haloplankton biomass differs slightly, presenting the gradual reduction from the interior part to the exterior part of the bay.

Concept and evaluation of bay health:the role of numerical model in the Yueqing Bay,China

To better evaluate the three-dimensional bay health and predict the dynamic bay health conditions, a concept of numerical bay health was introduced and a method of numerical bay health evaluation（NBHE） was developed.To support the NBHE method, a numerical bay health index（NBHI） system was constructed, which assess the natural and socio-economic effects on the entire bay. Five index groups are combined to formulate the NBHI,including geometry, hydrodynamics and sediment dynamics, bio-ecology, water quality and socio-economy.Each group has different number of indices selected and weighted using AHP method according to their importance. Data were mainly synthesized from a variety of numerical models together with monitoring programs, which provide superior to other approaches in discriminating data integrity and predicting data in future. The NBHE method using NBHI system was applied in the Yueqing Bay during spring tide in April 2007.According to the NBHE results, Sta. A, at the surface level of the estuarine mouth, has a healthy geometry condition, sub-healthy hydrodynamic and sediment dynamic condition, and unhealthy water quality and bioecology conditions. The integrated healthy score at Sta. A indicates its sub-healthy condition.

数据库模式的主动在线匹配方法

现有研究中常使用批量学习实现数据库间的模式自动匹配,但实际应用中模式匹配任务具有经验积累的增量性特征,难以为批量学习一次性标定大量学习样本。为此提出了一种数据库模式的增量式动态匹配方法,针对真实应用场景下模式匹配任务序贯到达经验动态增加的特点,采用集成学习与主动在线贝叶斯匹配算法增量学习数据库模式间的关联关系,在模式匹配样本渐进积累过程中,实现数据库模式的自动匹配。实验结果表明,给定相同初始训练样本,相比批量学习方法,所提方法能够将模式匹配F_(1)-score提高5%~33%。

Inference for Gompertz distribution under records

The parameter estimation is considered for the Gompertz distribution under frequensitst and Bayes approaches when records are available.Maximum likelihood estimators,exact and approximate confidence intervals are developed for the model parameters,and Bayes estimators of reliability performances are obtained under different losses based on a mixture of continuous and discrete priors.To investigate the performance of the proposed estimators,a record simulation algorithm is provided and a numerical study is presented by using Monte-Carlo simulation.

The quest for conditional independence in prospectivity modeling： weights-of-evidence, boost weights-of-evidence, and logistic regression

The objective of prospectivity modeling is prediction of the conditional probability of the presence T = 1 or absence T = 0 of a target T given favorable or prohibitive predictors B, or construction of a two classes {0,1} classification of T. A special case of logistic regression called weights-of-evidence （WofE） is geolo- gists＇ favorite method of prospectivity modeling due to its apparent simplicity. However, the numerical simplicity is deceiving as it is implied by the severe mathematical modeling assumption of joint conditional independence of all predictors given the target. General weights of evidence are explicitly introduced which are as simple to estimate as conventional weights, i.e., by counting, but do not require conditional independence. Complementary to the regres- sion view is the classification view on prospectivity modeling. Boosting is the construction of a strong classifier from a set of weak classifiers. From the regression point of view it is closely related to logistic regression. Boost weights-of-evidence （BoostWofE） was introduced into prospectivity modeling to counterbalance violations of the assumption of conditional independence even though relaxation of modeling assumptions with respect to weak classifiers was not the （initial） purpose of boosting. In the original publication of BoostWofE a fabricated dataset was used to ＂validate＂ this approach. Using the same fabricated dataset it is shown that BoostWofE cannot generally compensate lacking condi- tional independence whatever the consecutively proces- sing order of predictors. Thus the alleged features of BoostWofE are disproved by way of counterexamples, while theoretical findings are confirmed that logistic regression including interaction terms can exactly com- pensate violations of joint conditional independence if the predictors are indicators.

A New Hybrid Machine Learning Model for Short-Term Climate Prediction by Performing Classification Prediction and Regression Prediction Simultaneously

Machine learning methods are effective tools for improving short-term climate prediction.However,commonly used methods often carry out classification and regression prediction modeling separately and independently.Such a single modeling approach may obtain inconsistent prediction results in classification and regression and thus may not meet the needs of practical applications well.To address this issue,this study proposes a selective Naive Bayes ensemble model(SENB-EM)by introducing causal effect and voting strategy on Naive Bayes.The new model can not only screen effective predictors but also perform classification and regression prediction simultaneously.After being applied to the area prediction of summer western North Pacific subtropical high(WNPSH)from 2008 to 2021,it is found that the accuracy classification score(a metric to assess the overall classification prediction accuracy)and the time correlation coefficient(TCC)of SENB-EM can reach 1.0 and 0.81,respectively.After integrating the results of different models[including multiple linear regression ensemble model(MLR-EM),SENB-EM,and Chinese Multimodel Ensemble Prediction System(CMME)used by National Climate Center(NCC)]for 2017-2021,the TCC of the ensemble results of SENB-EM and CMME can reach 0.92(the highest result among them).This indicates that the prediction results of the summer WNPSH area provided by SENB-EM have a high reference value for the real-time prediction.It is worth noting that,except for the numerical prediction results,the SENB-EM model can also give the range of numerical prediction intervals and predictions for anomalous degrees of the WNPSH area,thus providing more reference information for meteorological forecasters.Overall,as a new hybrid machine learning model,the SENB-EM has a good prediction ability;the approach of performing classification prediction and regression prediction simultaneously through integration is informative to short-term climate prediction.

Personalized travel route recommendation using collaborative filtering based on GPS trajectories

Travelling is a critical component of daily life.With new technology,personalized travel route recommendations are possible and have become a new research area.A personalized travel route recommendation refers to plan an optimal travel route between two geographical locations,based on the road networks and users’travel preferences.In this paper,we define users’travel behaviours from their historical Global Positioning System(GPS)trajectories and propose two personalized travel route recommendation methods–collaborative travel route recommendation(CTRR)and an extended version of CTRR(CTRR+).Both methods consider users’personal travel preferences based on their historical GPS trajectories.In this paper,we first estimate users’travel behaviour frequencies by using collaborative filtering technique.A route with the maximum probability of a user’s travel behaviour is then generated based on the naïve Bayes model.The CTRR+method improves the performances of CTRR by taking into account cold start users and integrating distance with the user travel behaviour probability.This paper also conducts some case studies based on a real GPS trajectory data set from Beijing,China.The experimental results show that the proposed CTRR and CTRR+methods achieve better results for travel route recommendations compared with the shortest distance path method.

Prehospital Identification of Stroke Subtypes in Chinese Rural Areas

Background： Differentiating intracerebral hemorrhage （ICH） from cerebral infarction as early as possible is vital tbr the timely initiation of different treatments. This study developed an applicable model for the ambulance system to differentiate stroke subtypes. Methods： From 26,163 patients initially screened over 4 years, this study comprised 1989 consecutive patients with potential first-ever acute stroke with sudden onset of the focal neurological deficit, conscious or not, and given ambulance transport for admission to two county hospitals in Yutian County of Hebei Province. All the patients underwent cranial computed tomography （CT） or magnetic resonance imaging to confirm the final diagnosis based on stroke criteria. Correlation with stroke subtype clinical features was calculated and Bayes＇ discriminant model was applied to discriminate stroke subtypes. Results： Among the 1989 patients, 797,689, 109, and 394 received diagnoses of cerebral infarction, ICH, subarachnoid hemorrhage, and other forms of nonstroke, respectively. A history of atrial fibrillation, vomiting, and diabetes mellitus were associated with cerebral infarction, while vomiting, systolic blood pressure _〉180 mmHg, and age 〈65 years were more typical of ICH. For noncomatose stroke patients, Bayes＇ discriminant model for stroke subtype yielded a combination of multiple items that provided 72.3% agreement in the test model and 79.3% in the validation model; for comatose patients, corresponding agreement rates were 75.4% and 73.5%. Conclusions： The model herein presented, with multiple parameters, can predict stroke subtypes with acceptable sensitivity and specificity before CT scanning, either in alert or comatose patients. This may facilitate prehospital management for patients with stroke.