115 results about "Bayes' theorem" patented technology

A method and apparatus that provides an expert system for determining respiratory phase during ventilatory support of a subject. Discrete phase states are partitioned and prior probability functions and observed probability functions for each state are defined. The probability functions are based upon relative duration of each state as well as the flow characteristics of each state. These functions are combined to determine phase probabilities for each state using Bayes' theorem. The calculated probabilities for the states may then be compared to determine which state the subject is experiencing. A ventilator may then conform respiratory support in accordance with the most probable phase. To provide a learning feature, the probability functions may be adjusted during use to provide a more subject specific response that accounts for changing respiratory characteristics.

An improved method of selecting and planning the performance of various maintenance activities on a facility such as a nuclear power plant includes determining the net present value of a number of future net savings that are expected to result from performance of the maintenance activity at a given time, and selecting and planning the maintenance activities in a fashion that maximizes net present value. The method includes, for each of a number of components and a number of time periods, determining a change in the probability that a component will fail within a time period, with the change resulting from an assumption that a maintenance activity is performed. The change in probability is multiplied with the losses associated with a failure in order to determine a gross savings from which costs are subtracted to determine net savings. The probabilities of failure may be determined from a probability failure model that has been derived from multiple sets of failure data that are characterized by Weibull distributions and are mathematically combined according to Bayes' Theorem. The maintenance activities may also be optimized according to a number of budget figures. An apparatus for performing the method is also disclosed.

A fault location method includes steps: 1, establishing correlation matrix between fault pattern and tested item, 2, optimizing correlation matrix through fault location rate FDR and failure isolation rate FIR, 3, optimizing correlation matrix through maintenance data, 4, when single board gone wrong, according to optimized correlation matrix positioning fault device. Said invention utilizes Bayes' theorem to calculate single board fault probability and determining replacing which device according to probability to realize fault positioning. Said invented method has continuously modification positioning function.

A method for probabilistic fatigue life prediction using nondestructive testing data considering uncertainties from nondestructive examination (NDE) data and fatigue model parameters. The method utilizes uncertainty quantification models for detection, sizing, fatigue model parameters and inputs. A probability of detection model is developed based on a log-linear model coupling an actual flaw size with a nondestructive examination (NDE) reported size. A distribution of the actual flaw size is derived for both NDE data without flaw indications and NDE data with flaw indications by using probabilistic modeling and Bayes theorem. A turbine rotor example with real world NDE inspection data is presented to demonstrate the overall methodology.

The invention discloses a probability-based indoor location method. The method comprises the steps of establishing a fingerprint database on the basis of a received signal strength indication (RSSI) sequence and probability distribution of M reference points relative to N access points (AP), and location coordinates of M reference points; acquiring the RSSI sequence and probability distribution of at least one to-be-tested point relative to N AP, wherein Laplace calibration is performed on the probability of at least one to-be-tested point relative to each AP; on the basis of the Bayes theorem, matching the RSSI sequence and probability distribution of the j to-be-tested point of at least one to-be-tested point with the fingerprint database, so as to acquire the probability of the j to-be-tested point on each of the M reference points; and ranking the M probabilities from big to small so as to form a probability list, taking the reference points corresponding to the first k probabilities in the probability list, and computing the oriented coordinate of the j to-be-tested point. Through adoption of the method, the indoor location accuracy can be improved.

An improved approach for classifying portable executable files as malicious (malware) or benign (whiteware) is disclosed. The invention classifies portable executable files as malware or whiteware after using Bayes Theorem to evaluate each observable feature of each file with respect to other observable features of the same portable executable file with reference to statistical information gathered from repositories of known whiteware and malware files.

The invention discloses a Bayes-method-based spaceflight valve part fault diagnosis method in the machining process and belongs to the technical field of expert systems and fault diagnosis. The method includes the steps that first, a data control chart in the spaceflight valve part machining process is drawn, and the abnormal modes of the control chart are recognized according to a certain judgment rule; then, a relational network between the fault causes and the fault results in the spaceflight valve part machining process is described through FMEA and FTA, and an expert system knowledge base is built; finally, the set of main fault causes is acquired through importance analysis and the Bayes' theorem, and the list of causes which are most likely to cause the abnormal modes of top events is calculated. Through the spaceflight valve expert system fault diagnosis method, quality monitoring and fault diagnosis can be performed in the spaceflight valve part machining process containing a large number of data sets, the defects in the aspect of data standardization in an existing method are avoided, and meanwhile an expert diagnosis system for the special objects of spaceflight valves is developed.

The invention discloses a method for recommending context-aware Web service on the basis of weighted time-space effects. The method includes building weighted time decay models so as to find user sets with preference similar to preference of current users under condition that time decay effects are considered; acquiring user sets in contexts similar to current contexts of the users by the aid of location-aware similarity mining algorithms; building relation models of time-space correlation and user preference so as to obtain network service call records conforming to the current preference of the users; predicting QoS (quality of service) values of service by the aid of Bayes theorems on the basis of the obtained user sets with the preference similar to the preference of the current users under the condition that the time decay effects are considered, the user sets in the contexts similar to the current contexts of the users and the network service call records conforming to the current preference of the users so as to obtain the service with the highest degrees of satisfaction; evaluating outcome of the obtained service. The method has the advantage that the recommendation accuracy can be improved by the aid of the method.

The present invention provides a technology which precisely restores a distribution of original information by carrying out an iterative calculation based on a distribution of degraded information and a transfer function including a phase characteristic of a transfer system. A method according to the present invention restores the original information using the degraded information and the transfer function in the frequency space. The method according to the present invention considers a distribution of the degraded information and a distribution of the original information as distributions of probability density functions, and considers the transfer function as a probability density function of a conditional probability. The most probable distribution of the original information according to the distribution of the degraded information is acquired by the iterative calculation by means of relational equations based on the Bayes' theorem relating to the probability density functions.

The present invention provides a cancellation algorithm of bilayer weighted parallel interference based on the studing of cancellation algorithm of partial parallel interference and based on the weighted parallel interference cancellation algorithm of the Bayes theorem. The algorithm not only has the advantage of weighted algorithm based on the Bayes, theorem which can achieve the minimum cost in bit grade judgement, but also can make up the deviation to the client signal estimation on the statistic means at the advantage possessing in partial weighted algorithm. At the same time, comparing with the cancellation algorithm of weighted parallel interference based on the Bayes theorem, the algorithm has increased the gain a lot at the low signal to noise ratio without mush increasing amount of calculations.

The invention discloses a combined model water level prediction method based on similarity search. Water level of previous days related to a day to be predicted is confirmed to be sequences to be matched by utilizing correlation coefficient. A series of water level time sequences which are not similar to the sequences to be matched are searched from historical data on the basis of similarity search and eliminated from original time sequences and then act as training sets of a prediction model. The method mainly comprises data preprocessing, similarity search and a combined prediction model. Data preprocessing aims at filling gap data and restoring error data. According to similarity search, a series of time sequences which are not similar to the sequences to be matched are eliminated from the historical data of previous years by utilizing dynamic bending distance and the fixed slide window technology. The combined prediction model has two basic models: a BP neural network improved by an LM algorithm and a support vector machine, and proportion of the basic models in current prediction is dynamically adjusted by utilizing the Bayes theorem according to prediction performance of each basic model at the previous moment. The high-precision and real-time requirements required by flood prevention and disaster resistance can be realized.

The invention discloses a method for estimating an SOC of a power battery based on anti-outlier robust unscented Kalman filter, and belongs to the technical field of power batteries. The method comprises the following steps that a state and observation equation of the power battery is designed through the combination of a composite model method and an ampere-hour method, a model equation of the vehicle-mounted battery is determined, and a battery equivalence model is established; model parameters are identified, relevant parameters of the battery model observation equation are identified by means of a recursive least square method, the iteration frequency is identified with the system input amount as continuous excitation, and therefore a final result is converged and tends to be stable; an improved anti-outlier robust unscented Kalman filter algorithm is adopted for estimating the SOC of the battery. By means of the method, a measurement error model is corrected into a normalized contaminated normal distribution model, a posterior probability of the occurrence of outliers is calculated in combination with the Bayesian theorem to serve as a weighting coefficient for the self-adaptive adjustment to measure and predict related variances and gain matrices, and the problem of outlier interference can be effectively solved.

A face detection training method, a detection method and an apparatus thereof are disclosed. A face detector comprises a first classifier and a second classifier. The training method comprises the following steps of collecting face and non-face images as training sample sets; extracting a brightness comparison characteristic of each training sample in the training sample sets; according to the brightness comparison characteristic, acquiring a corresponding fern characteristic; using the fern characteristic and a Bayes theorem to carry out training and acquiring the first classifier; using the brightness comparison characteristic and a minimized weight mean square error criterion to carry out P-round training, and acquiring P decision tree classifiers, wherein the P is greater than or equal to 1 and the P is an integer; and cascading the P decision tree classifiers so as to form the second classifier. The face detector acquired through applying the training method is used to carry out face detection so that complexity and a calculated amount of face detection can be reduced and detection efficiency can be increased.

An improved approach for classifying computer files as malicious (malware) or benign (whiteware) is disclosed. The invention classifies any computer file as malware or whiteware after using Bayes Theorem to evaluate each observable feature of each file with respect to other observable features of the same computer file with reference to statistical information gathered from repositories of known whiteware and malware files.

The invention discloses a novel optic disc separation method and system. The method comprises the steps that S100, an angular point detection algorithm is utilized to detect points of interest aroundan optic disc in an eye fundus image, and convex hulls surrounding the points of interest are calculated to extract an optic disc region image; S200, according to a feature similarity between pixel points of the optic disc region image, the pixel points are grouped, and super pixels capable of replacing a large quantity of pixel expression image features are obtained; S300, prior probability distribution is calculated based on the convex hulls and the super pixels, color histograms are subjected to statistical analysis inside and outside the convex hulls respectively, and an observation likelihood probability is calculated; S400, the posterior probability that each pixel point f belongs to an optic disc region is calculated according to the Bayesian theorem, and a posterior probability distribution diagram is obtained; and S500, the optic disc is separated from the eye fundus image based on the posterior probability distribution diagram and through standard Hough Transformation circledetection. By the adoption of the separation strategy from rough to fine, precise separation of the optic disc from the eye fundus image is realized under a Bayesian model framework.

The invention belongs to the technical field of X-ray pulsar autonomous navigation and discloses an X-ray pulsar navigation TOA estimation method based on Bayes estimation. Under the condition that the overall trend of a photon counting rate accords with the Poisson distribution, an X-ray photon arrival time sequence can be modeled into a non-homogeneous Poisson process; the flow characteristics of PSR B0531 + 21 pulsars accord with the Poisson distribution, and a Poisson distribution signal model is established and divided into a time-frequency model and a frequency-stabilizing model; the frequency-stabilizing model of the photon sequence is selected to perform Fourier transform and then the frequency-stabilizing model is converted into a frequency domain to obtain a photon flow probability function expression with time delay estimation parameters; the photon flow probability function expression is converted into a likelihood function capable of calculating a time delay parameter by using a Bayes theorem for solving; and a Tool multi-mode nested sampling algorithm is calculated by Bayes estimation, iteration is carried out, and the parameter estimation value of the likelihood function is further calculated. The invention effectively improves the TOA estimation precision within the observation time and meets the future engineering development requirement of pulsar navigation.

The invention provides a network virtualization environment fault diagnosis method based on the relevance of symptoms and faults. According to the network virtualization environment fault diagnosis method, in two symptom-fault relations including the relation of the virtual network observation symptom-a virtual fault node and relation of an RVF-a physical fault node, the same diagnosis method is adopted twice; firstly, a suspicious virtual fault node set is selected and obtained from an observation symptom set in the virtual network based on the Bayes' theorem; secondly, an intersection set of the observation symptom set and a suspicious symptom set is obtained and used as an actual symptom set, wherein the suspicious symptom set is composed of all relevant symptoms of the suspicious virtual fault node set; a virtual fault node set is generated according to the actual symptom set and is used as a virtual diagnosis result, wherein the virtual diagnosis result can explain the generation of all the symptoms of the actual symptom set; NRVFs of virtual faults are excluded based on the same method, so that physical faults are obtained. By the adoption of the network virtualization environment fault diagnosis method based on the relevance of the symptoms and the faults, more accurate diagnosis of a network virtual environment can be achieved.

A method and apparatus that provides an expert system for determining respiratory phase during ventilatory support of a subject. Discrete phase states are partitioned and prior probability functions and observed probability functions for each state are defined. The probability functions are based upon relative duration of each state as well as the flow characteristics of each state. These functions are combined to determine phase probabilities for each state using Bayes' theorem. The calculated probabilities for the states may then be compared to determine which state the subject is experiencing. A ventilator may then conform respiratory support in accordance with the most probable phase. To provide a learning feature, the probability functions may be adjusted during use to provide a more subject specific response that accounts for changing respiratory characteristics.

The invention relates to a method for automatically filtering junk mails and a system thereof. The method for automatically filtering junk mails comprises the following steps: S1, a client receiving mails; S2, performing word segmentation operation on the mail content; S3, performing sample classification by using naive Bayes; S4, calculating the occurrence probability of each category under the occurrence condition of the item; if the mail is a normal mail, sending the mail to a client; and if the mail is a junk mail, cleaning the mail. According to the method, the probability probe of the Bayesian theorem is adopted for judgment. The types of the to-be-classified items are calculated by learning the mail word set samples through Bayesian classification, and the junk mail is finally intercepted according to calculation processing results, so that the effect is very obvious, the junk mail harassment frequency is greatly reduced, and the optimal user experience is provided for users.

The invention relates to an algorithm for correcting structure model parameters based on a frequency response function, which comprises the following steps of: acquiring time history data and time history response data, and introducing a multivariate circle symmetry proportion distribution theorem to derive and obtain a probability density function and a covariance matrix of the actually measuredfrequency response function; Introducing a prediction error and a to-be-corrected parameter to obtain a covariance matrix containing the to-be-corrected parameter; Obtaining a probability density function of a frequency response function under the action of single-point excitation according to the determinant and the inverse theorem of the matrix; Obtaining a maximum likelihood function expressedin a form of a maximum likelihood function and a logarithm maximum likelihood function according to a maximum likelihood principle; Obtaining a posterior probability density function of the random variable according to the Bayesian theorem; And expressing the posterior probability density function as a logarithm likelihood function form, so that an objective function is obtained. The uncertainty of the correction parameters is quantized, the calculation precision of the correction parameters is improved, and the correction of the structure finite element model is realized.

In magnetocardiography, an assisting technique for a doctor diagnosing a heart disease will be provided. In the magnetocardiography, magnetism to be generated from a subject's heart is measured by a magnetic sensor, and from this measured data, a position of center of gravity of excitement wave front in a cardiac muscle excitement propagation process and a transition vector (wave front vector) is calculated by an analyzing computer. With this wave front vector as a feature parameter, the Mahalanobis' distance is calculated from a data base obtained by diagnosing in the past, a possibility (probability) of the subject having a heart disease is presumed on the basis of the Bayes' theorem for displaying. Also, by superimposing on a heart picture, the wave front vector position is displayed, and a cluster classification position in the wave front vector is displayed. Also, by the similar retrieval, another subject extremely similar to the subject in the feature parameter will be retrieved from the data base for screen displaying.

The invention relates to a hybrid state space model-based sensor data blind correction method and belongs to the field of sensor networks. The method includes the following steps that: S1, a sensor data blind correction problem is modeled as an HSSM (hybrid state space model); S2, the a posteriori distribution of nonlinear observation signal parameters in the HSSM is obtained through using a UT-FBalgorithm according to relations between various parameters in the HSSM, and the a posteriori distribution of the other parameters in the HSSM is obtained through using Bayes' theorem and Dirichlet process; S3, the IMCMC (Iterative Markov Chain Monte Carlo) sampling method is used to iteratively collect the samples of the a posterior distribution of the parameters; and S4, some initial samples ofa sample set obtained by means of sampling in the S3 are discarded according the nature of the Markov chain, and the average value of an obtained sample set is solved, so that the estimated values ofcorrected parameter gain and offset are obtained. With the method of the invention adopted, the accuracy of sensor correction is improved under the premise of ensuring that the established model is close to the real application scenario of a sensor network.

The invention relates to the field of insurance informatization, and relates to an intelligent insurance recommendation method based on Bayesian. Bayesian algorithm is used to train the data after feature extraction, and the recommendation model is established. The recommendation model is imported into the online system of insurance products, and the user characteristic data is input to obtain theprobability of the user purchasing the insurance products as the recommendation result of the intelligent insurance. The invention also provides an intelligent insurance recommendation system based on Bayesian. The invention realizes intelligent recommendation of insurance products and improves user experience by analyzing and processing data, carrying out probability statistical calculation by using Bayesian theorem through historical sales data, training corresponding insurance recommendation model and applying it in online insurance system.

The invention provides a commodity price forecasting method and a device, which relate to the technical field of data analysis, and obtain target sample data of commodity price forecasting. Target sample data includes: training sample data; determining a prior distribution function of each parameter in the preset input-output relation function model based on the preset input-output relation function model; according to a priori distribution function, training sample data and Bayesian theorem, the commodity price forecast distribution function model is obtained. The new input parameter data isinputted into the commodity price forecast distribution function model, and the output result of the commodity price forecast distribution function model is calculated as the commodity forecast pricecorresponding to the new input parameter data. The invention establishes a commodity price prediction distribution function model based on the collected sample data of parameters related to commodityprice prediction, predicts the output result of new input data through the distribution function model, and improves the accuracy and credibility of the commodity price prediction result.

A service that leverages an innovatively extended version of Bayes' Theorem to provide comprehensive spam detection and optional spam elimination capabilities within established wireless messaging paradigms such as, possibly inter alia, Short Message Service, Multimedia Message Service, Wireless Application Protocol, and IP Multimedia Subsystem. The service may optionally leverage the capabilities of a centrally-located Messaging Inter-Carrier Vendor.

The present invention provides a technology which can perform a precise restoration to a distribution of original information by carrying out an iterative calculation using only a distribution of degraded information. The method according to the present invention performs the restoration to original information based only on degraded information. The method according to the present invention considers the distribution of the degraded information and the distribution of the original information as distributions of probability density functions, and considers a transfer function of a transfer system as a probability density function of a conditional probability. The most probable estimations of the distribution of the original information with respect to the distribution of the degraded information and the most probable estimations of the distribution of the transfer function are alternately carried out to finally restore the original information by means of the iterative calculation based on the Bayes' theorem for probability density functions.