67 results about "Probability assignment" patented technology

A classification system includes memory which stores, for each of a set of classes, a classifier model for assigning a class probability to a test sample from a target domain. The classifier model has been learned with training samples from the target domain and from at least one source domain. Each classifier model models the respective class as a mixture of components, the component mixture including a component for each source domain and a component for the target domain. Each component is a function of a distance between the test sample and a domain-specific class representation which is derived from the training samples of the respective domain that are labeled with the class, each of the components in the mixture being weighted by a respective mixture weight. Instructions, implemented by a processor, are provided for labeling the test sample based on the class probabilities assigned by the classifier models.

The invention discloses a hierarchical multi-source data fusion method for a pipeline linkage monitoring network, which comprises the following steps: carrying out data level preprocessing for various primary linkage detection signals acquired by a sensor at a common node of the monitoring network by using wavelet transformation, and extracting leakage-sensitive characteristic parameters; establishing a characteristic level data fusion model based on an ant colony neural network, processing the leakage characteristic parameters extracted by various sensors on the node, and constructing an elementary probability assignment function of evidence according to the output result of the ant colony neural network; and carrying out evidence synthesis at a cluster-head node according to an evidence combination rule, and making final decisions according to a maximum trust value method. The invention provides the hierarchical multi-source linkage detection data fusion method from the data level and characteristic level to decision level, and solves the multi-source data processing problem of the pipeline linkage monitoring network; and the method utilizes the linkage detection information acquired by various sensors in the network so as to effectively improve the accuracy rate of leakage identification.

The invention provides a robot failure diagnosis method achieved by multi-mode fusion inference. According to the method, results obtained from inference of various inference engines (such as inference engines based on rules, neural networks, the Bayesian network and evidence theories) are fused with the evidence theory method to obtain an inference result with higher credibility. The method mainly comprises the steps of determining an identification frame, converting the inference results of the various inference engines into elementary probability assignments, assigning a weight to each inference method, fusing the elementary probability assignments with the Dempster combination rule, and making decisions by means of gamble probability conversion. According to existing fusing inference methods, either a result obtained from last inference is used by next inference or different kinds of inference are adopted in different stages of a system, and therefore inference of the same information is only conducted with one inference method and the certainty can not be guaranteed. According to the robot failure diagnosis method achieved by multi-mode fusion inference, parallel inference is adopted, and the inference results obtained from the four inference methods are fused to improve the certainty of the inference result. The robot failure diagnosis method achieved by multi-mode fusion inference can be applied to failure diagnosis of multiple fields such as robots and can determine failures of a failure equipment system with a large number of uncertain factors.

An improved probabilistic lane assignment method for detected objects in the scene forward of a host vehicle. Road / lane model parameters, preferably including an angular orientation of the host vehicle in its lane, are estimated from host vehicle sensor systems, taking into account measurement uncertainty in each of the constituent parameters. A probabilistic assignment of the object's lane is then assessed based on the road / lane model parameters and object measurements, again taking into account measurement uncertainty in both the road / lane model and object measurements. According to a first embodiment, the probabilistic assignment is discrete in nature, indicating a confidence or degree-of-belief that the detected object resides in each of a number of lanes. According to a second embodiment, the probabilistic assignment is continuous in nature, providing a lateral separation distance between the host vehicle and the object, and a confidence or degree-of-belief in the lateral separation distance.

A multi-model dynamic soft measuring modeling method comprises the steps of establishing multiple sub models by utilizing a self-adaptive fuzzy core clustering method and a least square support vector machine; then taking a probability distribution function constructed by a proof synthesis rule as a weight factor to perform fusing on sub model output to obtain the output of multiple models; finally performing dynamic estimation on predicted errors of the multiple models by combining an autoregression moving average model.

A method for encoding and decoding a sequence is provided. The method comprises searching a set of candidate trees varying in size for a tree T having a plurality of states. Tree T provides a structure that relatively minimizes code length of the sequence from among all the candidate trees. The method further comprises encoding data conditioned on the tree T, which may be a generalized context tree (GCT), using a sequential probability assignment conditioned on the states of the tree T. This encoding may use finite state machine (FSM) closure of the tree. Also provided are methods for decoding an encoded binary string when the encoded string includes a full tree or generalized context tree, as well as decoding an encoded string using incomplete FSM closure, incremental FSM, and suffix tree construction concepts.

Methods and systems for processing records include extracting feature vectors from words in an unstructured portion of a record. The feature vectors are weighted based similarity to a topic vector from a structured portion of the record associated with the unstructured portion. The weighted feature vectors are classified using a machine learning model to determine respective probability vectors that assign a probability to each of a set of possible relations for each feature vector. Relations between entities are determined within the record based on the probability vectors. An action is performed responsive to the determined relations.

An apparatus capable of detecting location of object contained in image data and its detecting method are disclosed. The apparatus comprises an image capturing module, a weight assignment module, and a processing module. The image capturing module is for capturing an image. The weight assignment module performs the pixel weight / probability assignment according to the priori information and the image, and figures out the initial gravity center of the object according to the object location initialization. The processing module performs the statistical analysis according to the result of the pixel weight / probability assignment and the initial gravity center of the object so as to obtain the analysis result and update the object location. The processing module determines whether or not the analysis result meets the preset value, if it does, the processing module outputs an estimated result; if it doesn't, the processing module repeats the foregoing processes.

The invention relates to a fault tree diagnosis method for a typical fault of a reciprocating compressor coupled to a triangular fuzzy number. The method comprises the steps of (1) classifying variousfault types with each key component as a research object according to an operating mechanism of the reciprocating compressor, and establishing a fault tree of the reciprocating compressor typical fault, (2) establishing a minimum cut set of top-level events according to the logic relationship of upper and lower events of the fault tree, (3) determining the probability of occurrence of fault treetop-level events and each node layer event by combining the triangular fuzzy number, and (4) judging the location of a key component of a compressor fault according to measured data of characteristicparameters and according to the logic relationship and outputting a fault mode. According to the method, the probability that a triangular fuzzy function expresses a fault event is integrated into a fault tree diagnosis method, a multi-factor feature causing the fault is reflected, and probability assignment is allowed to have a certain degree of error. The method has the advantages of convenientapplication, high diagnostic accuracy and high reliability.

An improved system and method is provided for feature selection for text classification using subspace sampling. A text classifier generator may be provided for selecting a small set of features using subspace sampling from the corpus of training data to train a text classifier for using the small set of features for classification of texts. To select the small set of features, a subspace of features from the corpus of training data may be randomly sampled according to a probability distribution over the set of features where a probability may be assigned to each of the features that is proportional to the square of the Euclidean norms of the rows of left singular vectors of a matrix of the features representing the corpus of training texts. The small set of features may classify texts using only the relevant features among a very large number of training features.

The invention relates to an uncertain data provenance query processing method based on D-S evidence theory. The method comprises the following steps of: taking selection, projection and connection query operation related to an uncertain data table as a representative, acquiring elementary probability assignment of each input data item to a result data item from a provenance expression which describes SPJ query operation; based on an evidence combining rule in the D-S evidence theory, calculating the combined influence of the uncertainty of a plurality of input data items on the uncertainty of each result data item, and acquiring the probability assignment of each result data item; and performing standardization according to the probability assignment of each result data item, and calculating the belief value and the likelihood value of each result data item, so that the uncertainty of the result data item is determined, and if the uncertainty of the result data item accords with the result obtained on the basis of an input uncertain data-based probable world example, demonstration and evaluation can be performed on the basis of the pair of provenance query results.

The invention discloses an intelligent substation fault diagnosis method based on a probability weighting bipartite graph method. The method comprises the steps that intelligent substation symptom information is obtained, and interrogation information is obtained according to practical needs of fault analysis; all the information obtained in the first step is subjected to priority classifying, the symptom information and the interrogation information are subjected to probability assignment, and a probability weighting bipartite cause-and-effect graph of a practical system is obtained; and according to the probability weighting bipartite graph obtained in the second step, a fault locating algorithm based on the Bayes formula is used for computing the probability of each fault which may happen, final fault locating is carried out, a fault element is determined, and diagnosis results are output. According to the method, the fault of a primary system of an intelligent substation can be diagnosed, meanwhile the fault of a secondary system can be diagnosed, a diagnosis process is deliberate and reliable, the diagnosis results are accurate, and the great novel fault diagnosis method is provided for intelligent substation secondary system fault diagnosis.

A system, method and computer program product for allocating computing resources to process a plurality of data streams. A system for allocating resources to process a plurality of data streams. The system includes, but is not limited to: a memory device and a processor being connected to the memory device. The system receives at least one query from a user. The system obtains at least one sub-query associated with the at least one query. The system identifies at least one data stream associated with the at least one sub-query. The system computes at least one probability that the at least one sub-query is true. The system assigns the computing resources to process the data streams according to the computed probability.

Provided are a word latent topic estimation device and a word latent topic estimation method which are capable of hierarchically performing processing and which are capable of rapidly estimating latent topics of a word while taking into consideration a mixed state of topics. The present invention is provided with: a document data addition unit (11) which inputs a document which includes one or more words; a level setting unit (12) which sets a number of topics at each level in accordance with a hierarchical structure of topics for hierarchically estimating latent topics of a word; a higher-level constraint creation unit (15) which, on the basis of results of topic estimation at a given level with regard to a word within the document, creates a higher-level constraint indicating an identifier of a topic for which there is a possibility of being assigned to the word and a probability of being assigned to the topic; and a higher-level-constraint-attached topic estimation unit (13) which, when estimating the probability of each word being assigned to each topic, refers to the higher-order constraint, uses the probability of being assigned to a parent topic at the higher level as a weight, and performs estimation processing to a lower-level topic.

The invention discloses a farmland multi-source information dynamic adjustment fusion method and system, and the method comprises the steps: determining evidence factors and an identification framework, and calculating the probability distribution value of each evidence factor for each proposition in the identification framework; calculating a conflict coefficient according to a probability assignment value and a calculation formula of a conflict coefficient in a D-S evidence theory, and judges whether the conflict coefficient is in a set threshold interval, and if not, carrying out data fusion by using a classical D-S evidence theory synthesis rule, if so, adopting an average evidence factor to replace a probability distribution value of a conflict factor to correct an evidence source, and adopting a classic D-S evidence theory synthesis rule to carry out data fusion; or improving the classic D-S evidence theory synthesis rule according to the weight coefficient of each evidence factor and a historical accumulated data factor; and adopting the improved classic D-S evidence theory synthesis rule to perform data fusion, so that the reliability and reasonability of farmland monitoring data fusion are improved, and the decision risk is reduced.

The invention discloses a multi-attribute die-casting machine die pattern recognition method based on a D-S evidence theory. The method comprises the following steps that a target recognition problem is analyzed, and a systematic proposition set and a knowledge recognition framework are established for confirming multiple die-casting modes of a die-casting machine; aiming at a target information system, evidence bodies based on the recognition framework is established; characteristics of each proposition is combined based on known evidence body information, an elementary probability assignment of each evidence body is calculated by adopting the Hamming distance; based on the elementary probability assignment, a reliability section of each proposition in the recognition framework is calculated under the effect of each evidence body; by utilizing a D-S evidence composition rule, the elementary probability assignment and the reliability area are calculated under the combined effect of all the evidence bodies, a mode with the higher reliability is selected to be subjected to matching. According to few times of trying and technological modifications, a quality conformance casting is obtained. The production efficiency is greatly improved, and the production cost is lowered.

The invention discloses an evidence synthesis method and module, and a multiple Agent diagnosis system. The evidence synthesis method includes the steps: calculating the distance between every two evidences in an evidence set; according to the distance between every two evidences, calculating the similarity between every two evidences; according to the similarity between every two evidences, calculating the support degree of each evidence; according to the support degree of each evidence, calculating the credibility of each evidence; and taking the credibility of each evidence as weight, performing weighted average on the elementary probability assignment of the evidences in the evidence set, and synthesizing the evidences in the evidence set so as to solve the conflict problem of evidences and realize valid synthesis of multi-source evidences.

The invention relates to a marine site zone petroleum entrapment master control factor identification and favorable exploration target sorting method. The marine site zone petroleum entrapment mastercontrol factor identification and favorable exploration target sorting method is characterized in that as the marine site oil gas exploration degree and the geological cognition degree are low, through pairwise comparison of petroleum entrapment factors (source rock, reservoir, cover, entrapment, migration and storage) of one zone in a offshore petroliferous basin, assignment is performed according to the degree of importance of petroleum entrapment; a 6*6 identification matrix A of the zone petroleum entrapment master control factors is established, and the magnitude of the characteristic vector value of the 6*6 identification matrix A reflects the degree of importance of the six factors: source rock, reservoir, cover, entrapment, migration and storage in the zone for control of petroleumentrapment; through combination with comprehensive geological research, probability assignment exists in the source rock, reservoir, cover, entrapment, migration and storage of the exploration targetin the zone, the degree of importance of the six factors: source rock, reservoir, cover, entrapment, migration and storage in the zone can be weighed, and favorable exploration target comprehensive sorting on the zone is performed, and field exploration disposition and decision for a petroleum company can be provided.

The invention discloses a judgment method for a congestion event. The judgment method comprises the following steps of: fusing the congestion event probability of vehicles within the communication range with the credibility of the vehicles within the communication range according to the evidence integrating rule to obtain a global probability assignment function m(A) of congestion events and a global probability assignment function m(B) of non-congestion events, wherein the credibility magnitude of the vehicles is decided by the transmitting time and the transmitting distance when the vehicles transmit information and the physical qualities of vehicles; calculating the difference between the maximum waiting time and the minimum waiting time of the vehicles within the communication range; when the difference between the maximum waiting time and the minimum waiting time is greater than a minimum waiting threshold of the congestion and the difference between the global probability assignment function m(A) of the congestion events and the global probability assignment function m(B) of the non-congestion events is greater than or equal to the preset judgment threshold, determining the current events to be congestion events. According to the judgment method disclosed by the invention, the decision misjudgment due to short-time waiting events is avoided and the accuracy of judgment is improved.

The present invention discloses a method for designing a probability generator. According to a given probability distribution node, a Pow consensus mechanism of a blockchain is employed to calculate output nodes and takes the output nodes as an event to complete the design of the probability event generator. The method for designing a probability generator can realize the occurrence of the probability event according to the given probability.

The invention discloses an electronic device fault diagnosis method based on grey correlation analysis and improved DS reasoning, and relates to the technical field of fault diagnosis. The invention aims to solve the problem that fault diagnosis and fault prediction capabilities of node test equipment cannot be accurately realized by adopting a simple mode in the prior art. The method comprises: extracting fault characteristics of the tested electronic device to construct a parameter subspace of each fault characteristic; calculating the gray correlation degree of each fault characteristic parameter in the parameter subspace of each fault characteristic and the characteristic parameter in the fault type sample by adopting a gray correlation analysis method; selecting a fault correspondingto the characteristic parameter with the maximum correlation degree from the plurality of grey correlation degrees as a preliminary fault type of the tested electronic device; and performing probability assignment on each gray correlation degree to obtain a basic probability assignment of each fault type, and fusing the basic probability assignments by adopting an improved DS evidence reasoning fusion method to obtain a final fault type of the tested electronic device. The method is used for diagnosing electronic device faults.

The invention discloses a data fusion method and device, and relates to the technical field of computers. A specific embodiment of the method comprises the following steps: establishing an identification framework; Calculating a basic probability assignment, provided by any homogeneous data, of the to-be-identified object based on the identification framework by utilizing a preset basic probability distribution function; Performing orthogonal sum operation on the basic probability assignment provided by the homogeneous data acquired by any data acquisition system in a plurality of periods to obtain a time dimension probability assignment of the to-be-identified object in the data acquisition system; And performing orthogonal sum operation on the time dimension probability assignment of theto-be-identified object in the plurality of data acquisition systems to obtain the space-time dimension probability assignment of the to-be-identified object in the plurality of data acquisition systems and the plurality of periods, And determining an identification result of the to-be-identified object in the identification framework by utilizing the space-time dimension probability assignment.According to the embodiment, multi-source data can be effectively fused under the condition that the prior probability is not known, and a recognition result with high reliability is obtained.

The invention relates to a multi-information fusion fault diagnosis method and system for a thrust vectoring nozzle control system. The method comprises the steps of acquiring multi-source data information in a working state of the thrust vectoring nozzle control system, and dividing the multi-source data information into a training set and a test set; performing wavelet denoising processing on the multi-source data information in a training set to determine a multi-source information data set; performing fault feature extraction on each kind of multi-source data information in the multi-source information data set, and determining a fault feature parameter data set of a multi-information domain; performing dimension reduction processing on the fault feature parameter data set by adopting a principal component analysis method, and determining a main metadata set; and based on the main metadata set, fusing the multiple pieces of single evidence body information by utilizing a D-S evidence theory, determining a probability distribution value for each fault state, and identifying a fault type according to a maximum probability criterion. According to the invention, the accuracy of fault diagnosis can be improved.

The invention belongs to the field of multi-sensor high-conflict data fusion. The invention relates to a novel road condition state evaluation method based on JS divergence and a fuzzy evidence theory. The method aims at original conflict data collected by each sensor and comprises the following steps: firstly, reasonably allocating probability assignment of multiple subsets of focal elements of an uncertain part in probability to a single subset, wherein the JS divergence is used for measuring the distance sum of the same focal element in the evidence under different probability distributions; enabling the similarity coefficient to effectively measure conflicts between different evidence main bodies, then embedding a fuzzy reasoning mechanism is reasonably to objectively measure the conflict degree of the evidences, finally, using the support degree for obtaining the weight and weighting the evidences to obtain the average evidences, wherein the DS fusion rule of data fusion is used for fusing the average evidences for multiple times to obtain a reliable fusion result. The method has the beneficial effects that the road condition is reasonably described, the congestion condition and the change trend of the traffic road are specifically and quantitatively described by using the comprehensive value through the linear association of the comprehensive connection value and all connection components in the model, and the driving of a driver is assisted.

An adaptive traffic state estimation method contains the following specific steps: (1) in the fusion parameter training, an elementary probability assignment table of each data source is calculated according to traffic information of each data source; (2) in traffic state estimation of multisource fusion, the trained probability assignment tables of the sources are fused into a decision table, and a traffic state estimation result is obtained from the decision table according to real-time traffic information of each source; and (3) in adaptive adjustment, adaptive adjustment is carried out periodically on the probability assignment tables of the sources, and the adaptively-adjusted probability assignment tables of the sources are fused into a new decision table for real-time traffic state estimation during the next period. The probability assignment table of each data source can be adaptively adjusted periodically. Thus, data size and workload of the fusion parameter training at the earlier stage are minimized, and the adaptively-adjusted probability assignment tables can adapt to real-time urban road traffic state changing patterns. Traffic state estimation accuracy can be raised.

The invention belongs to the technical field of information processing, and discloses an unmanned aerial vehicle cluster target number detection method based on multi-sensor data fusion, comprising the following steps: step 1, two sensors of a radar and a communication signal detector respectively obtaining unmanned aerial vehicle cluster target number values, and obtaining basic probability assignment of detection data; step 2, fusing the elementary probability assignment in the step 1 by applying a D-S evidence theory fusion rule, and outputting the fused elementary probability assignment; and step 3, performing judgment decision on the fused basic probability assignment, and outputting an unmanned aerial vehicle cluster target number detection result. The D-S evidence theory-based unmanned aerial vehicle cluster target number detection method provided by the invention effectively solves the problem of low reliability of unmanned aerial vehicle cluster number detection by a single sensor, can provide more accurate data support for unmanned aerial vehicle cluster detection and defense, improves the accuracy of cluster defense effect evaluation, and a more reasonable defense decision can be made.

The invention discloses a hierarchical multi-source data fusion method for a pipeline linkage monitoring network, which comprises the following steps: carrying out data level preprocessing for various primary linkage detection signals acquired by a sensor at a common node of the monitoring network by using wavelet transformation, and extracting leakage-sensitive characteristic parameters; establishing a characteristic level data fusion model based on an ant colony neural network, processing the leakage characteristic parameters extracted by various sensors on the node, and constructing an elementary probability assignment function of evidence according to the output result of the ant colony neural network; and carrying out evidence synthesis at a cluster-head node according to an evidencecombination rule, and making final decisions according to a maximum trust value method. The invention provides the hierarchical multi-source linkage detection data fusion method from the data level and characteristic level to decision level, and solves the multi-source data processing problem of the pipeline linkage monitoring network; and the method utilizes the linkage detection information acquired by various sensors in the network so as to effectively improve the accuracy rate of leakage identification.

The invention relates to an uncertain data provenance query processing method based on D-S evidence theory. The method comprises the following steps of: taking selection, projection and connection query operation related to an uncertain data table as a representative, acquiring elementary probability assignment of each input data item to a result data item from a provenance expression which describes SPJ query operation; based on an evidence combining rule in the D-S evidence theory, calculating the combined influence of the uncertainty of a plurality of input data items on the uncertainty of each result data item, and acquiring the probability assignment of each result data item; and performing standardization according to the probability assignment of each result data item, and calculating the belief value and the likelihood value of each result data item, so that the uncertainty of the result data item is determined, and if the uncertainty of the result data item accords with the result obtained on the basis of an input uncertain data-based probable world example, demonstration and evaluation can be performed on the basis of the pair of provenance query results.

The invention discloses a greenhouse intelligent decision-making method based on a rough set theory and a D-S evidence theory, and relates to the technical field of intelligent agriculture, and the method comprises the steps: obtaining processed data through fuzzy C-means clustering processing, kernel solving and rough set attribute reduction; constructing a basic probability distribution function by using the rough set, and calculating the support degree among the influence factors of the greenhouse; and applying an improved D-S evidence theory, introducing the calculated BPA elementary probability assignment matrix, and constructing a confidence coefficient matrix to complete the combination of the greenhouse influence factors to obtain a decision result. A traditional SVM algorithm for small sample machine learning is utilized, a BPA basic probability assignment matrix is introduced, a decision result is obtained, and algorithm comparison verification is carried out on the decision result and a D-S evidence theory algorithm.

The invention discloses a weighted conflict evidence fusion method based on fuzzy classification, and the method comprises the following steps: firstly, obtaining the information of a plurality of evidences through a sensor, converting the evidences into single subset evidences through an interval probability assignment conversion formula, taking the evidences as vectors, obtaining the membership degree between the evidences, then, solving a fuzzy equivalence matrix through the membership matrix, and carrying out fuzzy classification on the evidence according to a threshold value; calculating the mutual support degree between the same type of evidences, obtaining the weight of each evidence in the same type of evidences based on the mutual support degree, and carrying out weighted average fusion on the evidences; and obtaining the fused new evidence, obtaining the weight of each evidence by combining the correlation support degree between the evidence and the credibility obtained by each evidence, and after the evidence is subjected to weighted fusion, carrying out secondary DS fusion. According to the method, fuzzy classification is carried out on the evidences, the weights of the evidences belonging to the same type and the evidences belonging to different types are obtained by adopting different methods, and the method has important theoretical significance and application value.