396 results about "Bayesian algorithm" patented technology

The invention discloses a method and a system for detecting a Chinese phishing website. The method comprises the following steps: S1, acquiring a URL (Uniform Resource Locator) by a client; S2, extracting URL features and website page content features as feature vectors respectively; S3, performing classification trainings on the feature vectors through an SVM (Support Vector Machine), an extended website page content feature NBC (Naive Bayes Classifier), a decision tree algorithm and a link and form processing method; S4, performing classification integration on classification training results, and judging whether a website with the URL is a phishing website. According to the method and the system, the URL features and the website page content features are extracted as the feature vectors, the corresponding classification trainings are performed through the SVM, the NBC, the decision tree algorithm and the corresponding processing of links and forms, prediction results are integrated by virtue of classification integration to obtain a final result, and therefore the classification accuracy is greatly improved.

The present invention is directed to systems and methods for isolating sentence boundaries between sentences in text. Sentences of the normalized document feeds or source text are separated by determining boundaries between individual sentences, by a Bayesian algorithm, that has been seeded with rule frequencies, developed from a previous training phase, that employed a text of sentences with marked boundaries between the sentences.

Foreground object matting uses flash / no-flash images pairs to obtain a flash-only image. A trimap is obtained from the flash-only image. A joint Bayesian algorithm uses the flash-only image, the trimap and one of the image of the scene taken without the flash or the image of the scene taken with the flash to generate a high quality matte that can be used to extract the foreground from the background.

The invention provides a Naive Bayesian classification based mobile phone spam short message filtering method and system. The system comprises a message intercepting module, a cache, a blacklist filtering module, a keyword filtering module and an intelligent Naive Bayesian classification filtering module. The message intercepting module is used for intercepting newly received short messages; the blacklist filtering module is used for filtering the new short messages according to a preset blacklist; the keyword filtering module is used for filtering the new short messages on the basis of preset keyword pairs; the intelligent Naive Bayesian classification filtering module is used for calculating probability that whether the new short messages are spam short messages or not by adopting a Naive Bayesian algorithm on the basis of a pre-trained feature word bank, and judging the new short messages as the spam short messages if the probability ratio exceeds a preset threshold, and as normal short messages otherwise. By the Naive Bayesian classification based mobile phone spam short message filtering method and system, through combination of the blacklist, the keywords, Naive Bayesian classification technology and Chinese word segmentation technology, the short messages are judged whether to be the spam short messages or not intelligently, so that the spam short messages are filtered.

The invention discloses a method and a system for track traffic failure recognition based on improved Bayesian algorithm. The method comprises the following steps of: 1) determining various failure modes and corresponding monitoring values of each traffic device according to circuit structure of the traffic device, and building a failure model aiming at each failure mode and corresponding monitoring value; 2) recognizing a parent child relation among the monitoring data according to the failure model, thus obtaining a standard failure sample data; 3) training with the standard failure sample data through a Bayesian algorithm to obtain a failure recognition model, wherein weight of a parent node in the failure recognition model of each failure mode is greater than that of a child node; 4) monitoring and acquiring various monitoring values of the traffic device in real time, and recording time sequence of the monitoring values; 5) recognizing data through the failure recognition model, and determining corresponding failure. By the method and the system, accuracy of failure recognition is improved, failure repair time is reduced, the device can perform failure self-diagnosis, and traffic safety is guaranteed in the operation and maintenance aspect and the device aspect.

The invention discloses a channel state information-based passive indoor positioning method. According to the method, ordinary equipment is utilized to build a data acquisition platform. The method specifically includes two stages, namely, an offline training stage and an online testing stage. According to the offline training stage, the channel state information data of each position where the body of a person is located, and the channel state information data are preprocessed, and then the preprocessed channel state information data are stored in a position fingerprint database, and a position-data fingerprint mapping relationship is established. According to the online testing stage, similarly, the data are preprocessed, a naive Bayes algorithm in machine learning is utilized to perform position classification. In order to further improve the accuracy of classification, a confidence method is introduced, and position misjudgment is decreased based on the classification result of a plurality of antenna pairs. With the method adopted, passive positioning of indoor people can be realized with low cost, and classification accuracy can achieve more than 90% under an optimal condition. The method of the invention has a certain application value in fields such as the intrusion detection field and the smart home field.

The invention relates to an old people fall detection and alarm system based on a 3D (3-dimensional) accelerometer and a gyroscope, belonging to the field of electronic information. The old people fall detection and alarm system is characterized in that a human body motion activity sensing module which is integrated with the 3D accelerometer, the gyroscope, a bluetooth chip and a microprocessor is embedded into a vest which is put on by old people, the 3D acceleration and the angular speed data of the activities of the old people are acquired in real time, the data of the activities is transmitted to a smart mobile phone in which fall detection software runs through bluetooth; the fall detection software automatically calculates the corresponding resultant acceleration alpha and the corresponding deflection angle theta according to the received activity data; and when the resultant acceleration alpha is larger than a resultant acceleration threshold alpha T and the deflection angle theta is larger than a deflection angle threshold theta T, a fall which occurs is judged, an alarm is given out by calling a designated contact person or transmitting a short message containing the position of the old people to the designated contact person through the mobile phone. The resultant acceleration threshold alpha T and the deflection angle threshold theta T used by the fall detection software are obtained by adopting a Bayesian algorithm. The fall detection and alarm system has the characteristics that the detection accuracy is high and the system is convenient for the old people to use.

The invention discloses an electricity larceny user identification method in combination with transformer area line loss and abnormal events. The electricity larceny user identification method comprises the steps of: acquiring transformer area and user data of at least one transformer area to be inspected, wherein the transformer area and user data comprises transformer area line loss data and basic data of each electricity utilization user in the transformer area; determining an abnormal line loss transformer area in which electricity larceny suspect users exist in a specified electricity utilization time period by applying a transformer area line loss anomaly detection method and the transformer area line loss data of the at least one transformer area to be inspected; determining a K-means clustering electricity larceny suspect user set, a support vector machine electricity larceny suspect user set and a Bayesian algorithm electricity larceny suspect user set for any abnormal line loss transformer area with the electricity larceny suspect users; and determining an electricity larceny suspect user list in the abnormal line loss transformer area after comprehensive evaluation. According to the electricity larceny user identification method, the identification rate of low-voltage electricity utilization abnormal users is effectively improved, and the workload of electricity utilization inspectors for investigation is effectively reduced.

Methods and apparatus are provided for estimating a location of a plurality of wireless terminals. Signal strength measurements are obtained for at least one packet transmitted by each of the wireless terminals; and a Bayesian algorithm is applied to the signal strength measurements to estimate the location of each wireless terminal. In an infrastructure-based deployment, signal strength measurements are obtained from one or more signal monitors. In a client-based model, signal strength measurements are obtained from a client associated with a respective wireless terminal.

The invention discloses a transmission line malfunction early-warning method and a system based on big data driving. The transmission line malfunction early-warning method comprises the following steps: acquiring related information of a power transmission line, wherein the related information of the power transmission line comprises electricity quantity, meteorological data, malfunction even sequence information and power grid topological data; extracting characteristics of the related information of the power transmission line, and furthermore establishing a malfunction factor mining databank and a malfunction judgment characteristic bank; based on a Naive Bayes algorithm, performing malfunction factor index mining on data in a current malfunction factor mining databank, and furthermore performing abnormal state tracking on an abnormal malfunction factor; transmitting the malfunction factor index of the abnormal malfunction factor to the malfunction judgment characteristic bank, matching an early-warning information time sequence generated in real time inside the malfunction judgment characteristic bank with a preset malfunction standard time sequence on the basis of a time sequence similarity malfunction matching method, outputting early-warning information corresponding to the matched early-warning information time sequences, and performing early-warning of malfunctions.

Provided is a mixing method for a brain-computer interface based on SSVEP and OSP. A subject wears an electrode cap. A SSVEP-OSP mixed paradigm is broadcast in front of the subject by means of a computer screen. The subject stares at any one of simulation units. By a collection system, an electroencephalogram signal generated when the subjects stares at a simulation target is magnified, filtered and subjected to analog-digital conversion by an electroencephalogram acquisition instrument. Digitized electroencephalogram data is inputted into a computer. An electroencephalogram signal feature extraction method based on a typical correlation analysis is adopted for extraction, classification and recognition of features of SSVEP. A support vector machine and naive bayesian algorithm are adopted for extraction and recognition of OSP features. A recognition result is displayed on the screen in order to feed back to the subject. Then neat recognition is carried out. The mixing method for the brain-computer interface based on SSVEP and OSP has following advantages: rate of information transmission of the method for the brain-computer interface is increased based on SSVEP; and the method is easy in operation, few in electrode number and many in target number.

The invention discloses a big data text classifying method based on cloud computing. The method comprises the following steps: respectively pre-processing training texts with class labels and without class labels to obtain corresponding training data sets; respectively carrying out feature selection on the training data sets to obtain corresponding dimensionally reduced training data sets; respectively calculating the dimensionally reduced training data sets according to a TFIDF weighted model, and respectively converting the training data sets to corresponding one-dimensional vectors; calculating the one-dimensional vectors with class labels according to a bayesian algorithm to obtain the prior probability of each class and the prior probability that each entry belongs to each class, and initializing the parameters of a bayesian classifier; utilizing an EM algorithm to optimize the parameters of the bayesian classifier so as to obtain a classifying model; carrying out text classification on the to-be-classified texts through the classifying model. Through combining a traditional naive bayesian classifying technology and Hadoop and EM algorithms, calculating speed limitation and training data limitation problems in actual application are improved, and the efficiency and the accuracy of the classifier are improved.

The invention discloses a Bayesian algorithm-based content filtering method. Content filtering is performed for text information in a 3rd generation mobile communication core network, text classification is performed by using a double threshold-based Bayesian algorithm, C1 is set to be normal information, C2 is set to be junk information, a classifier estimates the probability that a characteristic vector X which represents a data sample belongs to each class Ci, and a Bayesian formula for the estimation is that: P(Ci / X) = P(X / Ci) P(Ci) / P(X), wherein i is more than or equal to 1 and less than or equal to 2, the maximum value of a posterior probability is called the maximum posterior probability, for an error (a reference source is not found) of each class, the error (a reference source is not found) only needs to be calculated, a characteristic vector X of an unknown sample is assigned to the Ci class of the error (a reference source is not found) with the minimum risk value. Characteristic selection is performed by adopting document frequency (DF), and classification is performed by using minimum risk-based double threshold Bayesian decision. In a time division-synchronous code division multiple access (TD-SCDMA) mobile internet content monitoring system, the algorithm has higher controllability and can realize real-time high-efficiency classification of mass text information.

The invention relates to a mapping and positioning method based on information fusion, comprising the following steps: A. obtaining information of a robot body and surrounding environment by a sensor;B. processing data, and simultaneously creating a local map and performing feature matching to construct a global map; C. fusing data measured by an odometer and an IMU (Inertial Measurement Unit) byadopting a filtering algorithm to generate new pose information of a robot; at the same time, fusing environmental features acquired by a Kinect and a 2D laser radar based on a Bayesian algorithm toobtain new feature information; and finally, constructing a new local map based on the new environmental feature information and the new pose information; and D. performing feature matching on the newlocal map and the global map to complete data association, updating the global map by using the new local map, and outputting the global map. According to the mapping and positioning method of the invention, a raster map constructed based on the sensor information fusion is added with the environmental feature information, so that obstacles in a three-dimensional space can be detected; and meanwhile, the positioning precision is improved after fusion and complementation of the data of the odometer and the IMU.

Layer processing to grow a layer structure upon a substrate surface comprises supplying a vapor mixture stream to the substrate (28) to deposit constituents, monitoring growth with an ellipsometer (12) and using its output in real-time growth control of successive pseudo-layers. A Bayesian algorithm is used to predict a probability density function for pseudo-layer growth parameters from initial surface composition, growth conditions and associated growth probabilities therewith, the function comprising discrete samples. Weights are assigned to the samples representing occurrence likelihoods based on most recent sensor output. A subset of the samples is chosen with selection likelihood weighted in favor of samples with greater weights. The subset provides a subsequent predicted probability density function and associated pseudo-layer growth parameters for growth control, and becomes a predicted probability density function for a further iteration of pseudo-layer growth.

The invention relates to an emotion classification method and an emotion classification system. The classification method comprises the following steps: preprocessing the data of a sample to be tested; getting a feature word set of the sample to be tested; using a naive Bayes algorithm to perform calculation on the feature word set of the sample to be tested to generate the probability that the feature word set of the sample to be tested belongs to a category; and using a support vector machine to correct the probability that the feature word set of the sample to be tested belongs to a category, and determining the classification of the sample to be tested. According to the invention, Internet user emotion under the perspective of cross-cultural communication is analyzed in a more fine-grained way by building an emotional dictionary and emotion feature words in the field of cross-cultural communication, and the accuracy of emotion classification is improved.

The invention discloses a method and a device for Android malicious software classification based on Naive Bayes, wherein the method is used for extracting authority attribute characteristics and sensitive API (Application Programming Interface) attribute characteristics of malicious software mainly through a dynamic and static combination method, and performing classification through a Naive Bayes algorithm; and the device comprises a characteristic attribute set module, an authority characteristic attribute module, a sensitive API characteristic attribute module, a Naive Bayes classification model module, an output module and an updating module. The algorithm has firm mathematical foundation and steady classification efficiency, and needs relatively less parameters; the algorithm is simple; and malicious applications can be classified quickly and accurately, and thus, classification efficiency is improved.

The invention provides a health monitoring method and a health monitoring system, wherein the method includes the following steps: a mobile terminal sends a browsing request, a search request or a health report request to a server; when receiving the browsing request of the mobile terminal, the server sends a query instruction to a database according to the browsing request; when receiving the search request of the mobile terminal, the server sends a search instruction to the database according to the search request; when receiving the health report request, the server utilizes a naive Bayes algorithm to calculate the most relevant disease of a key term of the health report request according to the key term, and requests the information of the most relevant disease from the database; according to the received instruction, the database utilizes a web crawler to obtain data related to the instruction from the Internet; the server sends the data returned by the database to the mobile terminal according to a standard format. The health monitoring method and the health monitoring system which are provided by the invention provide query and health report functions, so users can conveniently query information at any time.

The invention provides a permanent resident population determination method and device and electronic equipment. The method comprises the following steps: acquiring signaling data; wherein the signaling data comprises user identifiers of a plurality of users, signaling interaction time of each user and identifiers of base stations for signaling interaction; determining residence data of each basestation based on the signaling data and a pre-acquired base station position table; determining area residence data of each area based on the residence data and preset area boundary data; and inputting the regional resident data into a permanent resident population model trained in advance based on a Bayesian algorithm to obtain permanent resident population of each region output by the permanentresident population model. In the mode, only a small amount of signaling data is needed for reckoning permanent resident population, a large amount of manpower does not need to be consumed to use a large amount of time for investigation and statistics, time cost and labor cost can be reduced, the implementation period can be shortened, dynamic change characteristics can be obtained, and the methodcan be used for cities with large population cardinal numbers and high floating population proportions.

The invention discloses a medical image disease classification method based on naive Bayes. According to an equipment type and image finding, diagnosis and other text information in a diagnosis report form, the disease type which an image examination result belongs to is automatically judged. Considering the influence of the independence assumption of naive Bayes classification in actual application, the method carries out disease clustering analysis by utilizing a K-Means clustering algorithm, data with the high similarity level are classified into the same cluster, data with the low similarity level are classified into different clusters, and meanwhile the number of disease categories is determined. The characteristics of high efficiency and high speed of a naive Bayes algorithm are utilized, classification precision is guaranteed, and meanwhile classification speed of medical image search is improved to a large degree.

The invention belongs to the technical field of wireless communications, and particularly relates to a method for reconstructing signals in an ultra-wide-band communication system by the adoption of a low-complexity improved Bayes compressed sensing algorithm. The method comprises the steps that firstly, sectioning and filtering are conducted on received signals, then a measuring matrix is used for conducting linear combination on the filtered signals again, after a series of relatively simple iterative operations, the expansion coefficient of original signals under the characteristic base can be measured, and therefore reconstruction of the original signals is realized. According to the method for reconstructing the signals, on one hand, the Bayes compressed sensing algorithm is improved by fully utilizing the characteristics of the characteristic base, the restorability is improved, more importantly, the complex matrix inversion process in the Bayes compressed sensing algorithm is avoided, and particularly, when the length of the signals is large and the order of the matrix is large, the signal restoration operation complexity can be effectively lowered.

The invention discloses a multilevel Email classification method based on Email content. For the problem that imbalance of samples has significant impact on an Email classification effect, a mode of establishing three-level classifiers is employed. According to the mode, a naive bayes algorithm, a support vector machine and a C4.5 algorithm combined voting mode is employed in establishment of a first level classifier; a random forest algorithm is employed in the establishment of a second level classifier; and a liblinear algorithm is employed in the establishment of a third level classifier. According to the method, for design of the classifiers, a multilevel classification mode is employed; levels of the algorithms are increased; through layer by layer filtering, under the condition of ensuring a recall rate, the accuracy of the classifiers is improved gradually; the balance of positive and negative samples is increased continuously; basic balance of the Email data is realized at the last level; the final classification effect is less influenced by the imbalance of the samples; and the relatively good effect is realized in Email classification.

The invention relates to television program recommendation. The invention provides a method and a system for recommending digital television program based on bayesian algorithm aiming that existing television program recommendation method only statistics user's viewing behavior but ignores playing behavior of television station. The method includes: collecting user's viewing recorder; analysing user's viewing recorder and obtaining user's viewing action parameter. The method also includes: obtaining television station playing action parameter according with viewing behavior parameter; predicting user's viewing probability to waiting-played television program according with viewing action parameter and playing action, and recommending television program to user thereof. The invention also provides a recommendation system corresponding to the method. Because combining the user's viewing action parameter and television playing action parameter when predicting user viewing special program, a recommendation result obtained by the television program recommendation is more accurately and reflects real user's viewing taste.

The invention belongs to the data processing of computer systems. As an important part of an emergency coordinating system, a warning situation automatic classification system shares database data with the emergency coordinating system. The core component thereof is a classifier which is a computer processing system, and the software constructed by Bayesian algorithm is used for carrying out the automatic classification of warning situation in the classifier. The process of construction includes two parts: the construction of the classifier and the operation of the classifier. The invention uses the theory of machine learning, takes the data of historical warning situation information as learning data and adopts the modified simple Bayesian algorithm to construct the classifier. New warning situation data is input to generate standard warning situation vector information in classification, and the category of the warning information can be determined by a composite probability formula according to a preset probability threshold value and the occurrence number of the TOKEN serial. The classification system not only is adaptable to various emergency systems, but also can be an independent system. The classification is rapid, accurate, normalized and scientific.

The invention particularly relates to a predication model establishing method based on a naive Bayesian algorithm. The predication model establishing method based on the naive Bayesian algorithm comprises the steps of collecting factor data related with cerebral apoplexy, converting qualitative data to quantitative data, calculating a correlation coefficient between each feature and a target valueby means of a Pearson correlation coefficient method, and quantitatively converting the correlation coefficient number to the weight; respectively processing discrete feature data and continuous feature data according to a traditional algorithm, a polynomial and a Gaussian model, lifting the influence of important features to a predication result by means of a weighted feature analysis method, and furthermore obtaining the predication model with high predication accuracy. According to the predication model establishing method based on the naive Bayesian algorithm, the predication model with high predication accuracy is finally obtained by means of a hybrid predication model, a feature weighting method, a sliding defining factor and a plurality of evaluation indexes; reference data can besupplied for clear diagnosis and treatment of a doctor; and furthermore an important meaning is realized for development of a national health industry.

The invention provides an Android malware detection method based on an improved Bayesian algorithm. The feature attributes of Android malicious programs and well-behaved programs are analyzed and classified through the improved Bayesian algorithm to realize the malware detection method based on the improved Bayesian algorithm. A judgment on whether software is malware is implemented from the aspect of permission application of applications. According to the method, a permission request label in an Android permission request mechanism is taken as a detection data source. The malware and well-behaved software are distinguished in a permission request label combination way, and a detection model is built by using the improved Bayesian algorithm. The improved Bayesian algorithm is characterized in that mutual independence among attributes of the data source is considered, and a naive Bayesian classifier is used for performing data modeling, so that the detection index is increased greatly, the detection accuracy is increased, and the false alarm rate is lowered.

The invention discloses a case text classification method based on naive Bayes, a system and a storage medium. The method comprises the following steps: a classifier is constructed based on the naiveBayes algorithm; Obtaining training samples to train the classifier; Obtaining text to be categorized; The text vector of the text to be classified is obtained by preprocessing the text to be classified; Inputting the text vector into a classifier, and calculating a posterior probability of the text to be classified belonging to each category according to a priori probability of each category anda priori probability of each feature word belonging to each category; The class with the highest posterior probability is output as the classification result. The invention fully considers the difference of the sample quantity between different categories, and takes the prior probability of each category and the prior probability of each characteristic word belonging to each category as the calculation factor of the classifier, so that the invention has better classification effect on case text classification. The invention can be widely applied in the field of data mining.

The invention belongs to the technical field of wireless communication and specifically relates to a phase noise compensation and suppression method for a large-scale MIMO system. The method providedby the invention mainly comprises the steps of carrying out maximal-ratio combining on receiving signals of antennas under the condition of assuming that phase noise is zero, thereby obtaining an initial value of a data symbol; carrying out iteration according to the obtained initial value through adoption of a variational Bayesian algorithm, wherein an estimation value of the data symbol is converged to a stable value under the condition that the receiving signal ri is known. The method has the beneficial effects that the data symbol can be judged in a high-order modulation mode within the large-scale MIMO system; and according to a judgment method, the inversion operation of a high-dimensional matrix is avoided in a traditional variational Bayesian inference algorithm, and the inversionoperation of the high-dimensional matrix is converted into the inversion operation of a diagonal matrix, so the system complexity is reduced, and the system performance is improved.

The invention discloses a method for filtering spam short messages based on user selection. The method comprises the following steps: (1) a user installs a client on a mobile phone and registers in a server to obtain an activation code after finishing the installation. The server establishes an account for each registered user, and the user logs in the server to set categories not needing to be filtered and to add security keywords; (2) a server end generates filtering keywords according to short messages in the latest spam short message library and non-spam short message library in a certain period by using Bayesian algorithm; (3) the client updates and downloads the filtering keywords from the server; the server accesses the account setting of the user before providing the download of the keywords, and then provides the filtering keywords to be downloaded to the client after tailoring according to the setting of the user; and (4) the client performs the filtration of the spam short messages according to a black and white list and the filtering keywords.