78 results about "Computational probability" patented technology

A multi-user turbo decoder combining multi-user detection and forward error correction decoding is disclosed that utilizes iterative decoding of received, interfering signals, and the construction of a decoding tree of the decoder is changed for each iteration of the decoding based on the previous conditional probability estimates of the value of the data bits of each signal making up the received, interfering signals. Before each iteration of multi-user decoding, a probability estimate is calculated that the value of the bit in a signal has a certain value for all of the data bits. Using the probability estimate a new decoding tree is constructed before each iteration of decoding such that the signal bit having the most reliable estimate is assigned to the lowest or root level of the tree. Using the probability estimate for the other signal bits, the signal bit having the next most reliable estimate is assigned to the second level of the tree, and so forth, with the signal bit having the least reliable estimate being assigned to the highest level of the tree adjacent the terminating nodes or leaves of the tree. By building the decoding tree in this manner for each iteration of symbol decoding, a reduced complexity search is more likely to include paths (and nodes) in the tree containing the correct value for the channel symbols.

Some embodiments set forth probability based caching, whereby a probability value determines in part whether content identified by an incoming request should be cached or not. Some embodiments further set forth probability based eviction, whereby a probability value determines in part whether cached content should be evicted from the cache. Selection of the content for possible eviction can be based on recency and / or frequency of the content being requested. The probability values can be configured manually or automatically. Automatic configuration involves using a function to compute the probability values. In such scenarios, the probability values can be computed as a function of any of fairness, cost, content size, and content type as some examples.

The present invention is directed to providing a spatial-multiplexed signal detection method that can improve the characteristics of spatial and temporal iterative decoding that is based on turbo principles. According to the method, when implementing factorization of conditional probability referred to as “likelihood” such that the conditional probability can be represented by the product of a plurality of conditional probabilities, the conditional probability being obtained for a received signal sequence in a spatial and temporal iterative decoding configuration based on turbo principles of soft-input soft-output detector 1 and soft-input soft-output decoder 2, the conditional probability for which factorization is possible is divided into a plurality of groups. When calculating this likelihood, the ordering among groups in which probabilities are calculated can be ordered such that groups that contain events that serve as the conditions of conditional probabilities in the groups are processed earlier. When calculating the probabilities in the groups, a metric operation method is used that uses semi-rings for estimating transmission sequences by means of the ratio of likelihoods of two exclusive events.

The invention discloses a medical terminology standardization method and system based on probability transfer matrix, which are designed for realizing mapping of universal short text (abbreviation, misspelling, daily expression, etc.) to medical standard terminology in the medical field. The medical terminology standardization method based on the probability transfer matrix comprises the followingsteps: constructing a medical terminology database; performing medical word segmentation and part-of-speech tagging; constructing a term-based probability transfer matrix framework; constructing a word vector model; calculating a probability matrix; calculating a probability of terms to be matched. The invention can realize the quick, efficient and accurate mapping of various diseases in the medical field corresponding to the ICD10 standard coding.

The invention discloses a method for determining arithmetic encoding probability interval value based on JPEG (Joint Photographic Experts Group) 2000 standard. The problems to be mainly solved are that the prior method has low processing speed and a plurality of interior storage bits and cannot effectively update the probability interval values of any double text symbol pairs. The determination method comprises the following steps of: firstly, judging the types of the double text symbol pairs; secondly, acquiring the probability value of the corresponding small probability symbol from the probability value estimation table, respectively updating the probability interval values of the different types of the double text symbol pairs, and meanwhile, calculating the shift counting value of the probability interval value; and finally, outputting the updated probability interval value. The invention has the advantages of high processing speed and few interior storage bits and can be applied to the image compression encoding of various high-speed digital devices, in particular to the encoding of the high-speed real-time satellite remote image.

The invention discloses a power system probabilistic-optimal power flow calculation method based on a stacked denoising autoencoder. The calculation method comprises the following main steps that: 1)establishing a SDAE (stacked denoising autoencoder) optimal power flow model; 2) obtaining the input sample X of a SDAE optimal power flow model input layer; 3) initializing the SDAE optimal power flow model; 4) training the SDAE optimal power flow model so as to obtain a trained SDAE optimal power flow model; 5) adopting a MCS (Modulating Control System) method to carry out sampling on the randomvariable of a power system to be subjected to probabilistic power flow calculation so as to obtain a calculation sample; 6) inputting training sample data obtained in S5 into the SDAE optimal power flow model which finishes being trained in S4) in one time so as to calculate an optimal power flow online probability; and 7) analyzing the optimal power flow online probability, i.e., drawing the probability density curve of the output variable of the SDAE optimal power flow model. The method can be widely applied to the probabilistic-optimal power flow solving of the power system, and is especially suitable for an online analysis situation that system uncertainty is enhanced due to high new energy permeability.

System and methods are provided for a supervised point map matcher. The supervised point map matcher learns parameters from historical data that provide insight into the optimal probabilistic metrics that inform the bias of probes heading and distance for segments on the roadway. Probability weights for segments are generated. A more accurate path based map matching algorithm is used to identify direction and heading errors in the historical probe data. Values for the probability weights are calculated using kernel density estimation and a gaussian probability density function. The probability weights are used to improve the real time performance of the point map matcher. A confidence value is calculated as a function of the probability weights and provided with the map matched results.

A method of designing a transport network having a plurality of network elements and a plurality of connections between the network elements by (a) defining a first network configuration and at least one alternative network configuration for the same transport network; (b) calculating for each network configuration, a probability function representing, for each maximum number of routable flows, the probability of routing such a number of flows in the network configuration currently considered; (c) calculating for each network configuration, a unit-cost-per-flow function calculated as the ratio between a sum of the costs relative to the network elements of the network configuration currently considered and the probability function; and (d) comparing the unit-cost-per-flow functions of the network configurations considered, for choosing a network configuration having a lowest unit-cost-per-flow value.

A quality control system in combination with an ECG data analysis system to analyze ECG data acquired by an ambulatory electrocardiography device via a cable having a plurality of leads connected to a subject is provided. The quality control system includes a memory having programmable instructions for execution by a processor to perform the steps of calculating a trend of a quality score in the ECG data dependent on a noise content in the ECG data; and calculating a probability that one of a hardware and the Hookup personal, each associated with collecting the ECG data from the subject, is a substantial cause of the quality score for the ECG data. The probability can be calculated based on a comparison the trend of the quality score associated with the Hookup personnel versus the trend of the quality score of the hardware employed in acquiring the ECG data.

Method for judging age brackets of users including acquiring consumption data of users and establishing models based on the consumption data. Establishing the models includes dividing the consumption data into training data and test data, calculating a number of the users of the training data in predetermined age brackets, calculating a number of each tertiary category of the training data in the predetermined age brackets, calculating probabilities that each tuple of the test data belongs to each of the predetermined age brackets based on the number of the users and the number of the tertiary categories, selecting the age bracket with the maximum probability as the age bracket to which the user corresponding to the tuple belongs, comparing errors between the predetermined age brackets and the selected age bracket to obtain a predictive error rate, and outputting models with predictive error rates larger than or equal to a predetermined threshold.

A quene management method for providing proportional drop rate used in the message forwarding system is characterized by using "weighted simple self-adaptive proportion controller" algorithm in a process of acquisition of transcient quene length to calculate message drop or label probability. The said algorithm is composed of two parts: 1 a fundamental algorithm of using quene controller to calculate the probobilty of message and in line with its kind to use related probability to drop or label the message through the gate; 2 with the adaptive algorithm, adjusting dynamically the parameter avg p namely the average value of the probability used in the fundamental algorithm. The experiment on a PC Intel P 111800 with network analog ns-2 has testified that it's easy to be realized with low cost is computing, simple parameter set, broad adaptability and automatically fit the variation of the network environment without changing parameter set.

A method and apparatus are provided for ranking documents according to relevancy scoring. In one implementation, a computer-implemented method is provided for receiving, from a database over a network, a document resulting from a search on a database, the document containing terms that match the search criteria. The method may calculate a standard deviation of a probability distribution function representing a distribution of the terms in the document that match the search criteria. The method may further determine relative distances between the terms in the document that match the search criteria according to the standard deviation. The method may further calculate a proximity boost value using the relative distances, and apply the proximity boost value to a base relevancy score of the document to determine a relevancy ranking. The document may then be ranked according to the relevancy ranking.

The invention relates to a recommendation algorithm based on adversarial learning and a bidirectional long-short-term memory network, which comprises the following steps of: 1, predefining a symbol, including A1) defining a heterogeneous information network, A2) defining a path in the heterogeneous information network, A3) in the heterogeneous information network G, defining a node connection sequence from a user u to an article i as a path, and defining that p = [v1, v2,..., vl], and p belongs to P; and 2, modeling as following: S1, modeling an embedded layer, and representing the embedded layer by using an initialized node vector; S2, constructing a sequence modeling layer, using the vector representation obtained through initialization in the step S1 as input and applying the input to an existing bidirectional LSTM model based on an attention mechanism to optimize vector representation of the node, and learning a coefficient matrix and an offset vector in the model; S3, setting a prediction layer, and finally calculating the probability; and S4, constructing an adversarial learning model. According to the method, the problem of node relation noise in the heterogeneous network isrelieved by learning the adversarial regularization item, adding the adversarial regularization item into a loss function and optimizing the model, the robustness of node embedding is improved, and the recommendation accuracy is ensured.

The invention discloses a random optimal trend calculation method based on a random response surface method. The influences of randomness of input variables are taken into consideration in an optimization process, and finally, a group of optimal solutions satisfying certain opportunity constraints are obtained. The method comprises the following steps: first of all, power system information is input, random input variables (disturbance variables) in a system are determined, the input variables are substituted as expected values, and an optimal scheduling scheme is obtained by performing deterministic optimal trend calculation by use of a primal-dual interior-point method; then, a probability trend is calculated by use of the random response surface method, a probability distribution of system state variables under the scheduling scheme is obtained, and random variables with correlation are processed by use of Nataf transformation; and finally, by use of a probability distribution function, whether the state variables satisfy restrictions of the opportunity constraints is determined, if not, upper and lower limits of the opportunity constraints are adjusted, and the steps of the deterministic optimal trend calculation and opportunity constraint examination are restarted until a scheduling scheme satisfying the opportunity constraints is obtained.

The present invention provides a method of performing an entropy decoding for a video signal including obtaining a context model initial value for a current slice; calculating a probability value based on syntax statistics of a previous slice; deriving weighted values corresponding to the context model initial value for the current slice and the syntax statistics of the previous slice; and updating the context model initial value for the current slice using the weighted values.

A method and apparatus are provided for ranking documents according to relevancy scoring. In one implementation, a computer-implemented method is provided for receiving, from a database over a network, a document resulting from a search on a database, the document containing terms that match the search criteria. The method may calculate a standard deviation of a probability distribution function representing a distribution of the terms in the document that match the search criteria. The method may further determine relative distances between the terms in the document that match the search criteria according to the standard deviation. The method may further calculate a proximity boost value using the relative distances, and apply the proximity boost value to a base relevancy score of the document to determine a relevancy ranking. The document may then be ranked according to the relevancy ranking.

System and method for balancing a number of incoming connections of nodes connected to a peer-to-peer network. The method includes receiving at a selecting node information about out-finger candidates. The information includes at least a quantifiable parameter for the out-finger candidates and the out-finger candidates are possible nodes to be connected to the selecting node. The method includes calculating in the selecting node, for each node of the out-finger candidates, a probability value by using a predetermined probability function that uses as input the corresponding quantifiable parameter; and selecting one node from the out-finger candidates to be an out-finger node for the selection node based on the probability value. The probability value is associated with a number of in-finger connections to the selected out-finger node.

A device and method for assessing the damage state of solid materials and structures subjected to loading. The device includes multiple AE sensors connected to the switch controller / Amplifier / AD convertor, the event sorting module, the spectrum assignment unit, the probability space resolver, the trajectory of damage state generator, power source, and a visual display. The method includes means to assess and analyze performance of solid materials and structures that accounts for the influence of microscopic random damage events statistically, the method including the steps of sorting the electric signals into a series of non-overlapping AE events; determining the spectra of the sorted events; computing the probability distribution of the spectra; computing the probabilistic entropy of the probabilistic distribution; and generating the trajectory of damage state.

At least one of a probability density for a video and a granularity mapping for the video based upon the probability density are generated at an analytics server. The probability density is calculated according to a predetermined set of points in the video and a pseudo-randomly generated set of points in the video. The probability density is non-degenerate. The granularity mapping for the video provides a mapping of the probability density to the video. The at least one of the probability density and the granularity mapping are sent from the analytics server to a video content server so that the video content server sends the video and the at least one of the probability density, and the granularity mapping to each of a plurality of clients. A set of beacons is generated by the each of the plurality of the clients according to the probability density and the granularity mapping.

The present disclosure relates generally to a calculated probability that an income value has been misrepresented in a risk analysis system. For example, the system may apply first data to a first machine learning (ML) model to determine a conservative income prediction associated with the data and apply second data to a second ML model to determine a probability that an overstatement of the income value would result in a change in an approval determination.

A model generation device for life prediction includes: an actual operation information generation unit that generates actual operation information indicating a relationship between a use time and a reliability of an object whose life is predicted, based on failure history information of the object by using an order-statistic calculation method; a probability distribution model generation unit that sets a number of division by which the use time is divided into periods, and then generates a probability distribution model that approximates the actual operation information for each of the periods obtained by dividing the use time; a calculation unit that calculates a goodness of fit of the probability distribution model to the actual operation information for each of the number of division by using an information criterion; and a determination unit that determines the probability distribution model at the number of division providing the highest goodness of fit.

The invention discloses a method and a system for dynamic demand response of a comprehensive energy system. The method comprises the following steps: acquiring outdoor temperature and illumination data, and generating a distribution function; solving a joint distribution function considering temperature and illumination correlation; solving a joint distribution function considering the temperatureand illumination dynamic correlation; obtaining an edge distribution function of each variable, and generating an N * K high-dimensional sample; performing scene reduction and dimension reduction processing on the high-dimensional sample; calculating the active power output in the distributed photovoltaic power supply system and the thermal load power in the thermal load; establishing a node cold / heat / electricity / gas load physical model based on multi-task learning, and calculating the comprehensive trend of electricity_heat_ gas; establishing a random response surface regression model basedon statistical machine learning, and quickly calculating the digital characteristics of the probabilistic load flow; and establishing a comprehensive energy system dynamic demand response optimizationmodel based on probabilistic load flow, and formulating a user side dynamic demand response strategy matched with photovoltaic power generation characteristics according to a solving result.

The invention discloses a reinforcement learning method and a related device, a regularization item used by the reinforcement learning method is a regularization item based on a sample, the regularization item based on the sample comprises additional rewards for executing different actions by a target intelligent agent, the availability degree of exploration behaviors is defined, mutual influence between executed actions is modeled, a probability density function does not need to be calculated when the regularization term based on the sample is used for reinforcement learning, the reinforcement learning efficiency can be improved, and particularly the reinforcement learning efficiency for complex strategies is improved. In addition, the sample-based regularization term is used for reinforcement learning, so that geometrical information of an action space can be utilized, and the method can be compatible with a wider strategy structure.