649 results about "Conditional probability" patented technology

A controller for a building management system is configured to analyze faults in the building management system. The controller detects a fault in the building management system by evaluating data of building management system using a system of rules. The controller determines a conditional probability for each of a plurality of possible fault causes given the detected fault. The controller determines the most likely fault cause by comparing the determined probabilities and electronically reports the most likely fault cause.

A method for diagnosis of a system made up of a plurality of interlinked modules includes receiving an alarm from the system indicative of a fault in one of the modules. Responsive to the alarm, a causal network is constructed associating the fault with malfunctions in one or more of the modules that may have led to the fault and relating a conditional probability of the fault to respective probabilities of the malfunctions. Based on the alarm and the causal network, at least one of the probabilities of the malfunctions is updated. A diagnosis of the alarm is proposed responsive to the updated probabilities.

A document annotation method includes modeling data elements of an input document and dependencies between the data elements as a dependency network. Static features of at least some of the data elements are defined, each expressing a relationship between a characteristic of the data element and its label. Dynamic features are defined which define links between an element and labels of the element and of a second element. Parameters of a collective probabilistic model for the document are learned, each expressing a conditional probability that a first data element should be labeled with information derived from a label of a neighbor data element linked to the first data element by a dynamic feature. The learning includes decomposing a globally trained model into a set of local learning models. The local learning models each employ static features to generate estimations of the neighbor element labels for at least one of the data elements.

In a wireless communication network, the location of an addressable receiver relative to the locations of a plurality of addressable sources of electromagnetic radiation is found using probabilistic models of the signal strength measured at the addressable receiver. The inventive method provides location determination on a finer spatial scale than was heretofore available. A region of interest is calibrated via a discrete-space radio map storing probability distributions of received signal strength at the measurement locations. The stored probability distributions are compensated for temporal variability and biases, such as through temporal correlations of the sampled received signal strength. A measurement of the signal strength at the addressable receiver from each of the plurality of addressable sources is used in conjunction with the discrete-space radio map to identify one of the coordinates thereof that maximizes the conditional probability P(x|s), where x is the radio map location and s is a vector of measured signal strengths. Small spatial scale variability can be compensated for using a perturbation technique. The method further implements a continuous-space estimator to return an estimated user location that falls between discrete-space radio map locations.

Methods of and systems for resolving multiple location estimate conflicts in a WLAN-positioning system are provided. Disclosed are methods to quantify the probability that a particular location estimate of a mobile device made by a Wi-Fi based positioning system is correct to within an arbitrary accuracy. Implementations use observed access point cluster size, age information for access point location determination, and/or cumulative distribution functions that characterize the conditional probability that one or more access points detected by the mobile device have relocated within a specified time interval to make the probability determinations.

The invention relates to speech emotion recognition equipment and a speech emotion recognition method. The speech emotion recognition equipment comprises an analysis device, a probability calculating device and an emotion judging device, wherein the analysis device matches emotional characteristics of input speech with a plurality of emotion models so as to ensure a plurality of possible emotional states; the probability calculating device calculates the final probability of the possible emotional states in the previous emotional states of a speaker according to the knowledge of conditional probability converted among the emotional states in the speaking process of the speaker; and the emotion judging device selects a possible emotional state with the maximum final probability from the possible emotional states as the emotional state of the input speech.

The invention relates to a voice-and-facial-expression-based identification method for dual-modal emotion fusion. The method comprises: S1, audio data and video data of a to-be-identified object are obtained; S2, a face expression image is extracted from the video data and segmentation of an eye region, a nose region, and a mouth region is carried out; S3, a facial expression feature in each regional image is extracted from images of the three regions; S4, PCA analysis and dimensionality reduction is carried out on voice emotion features and the facial expression features; and S5, naive Bayesian emotion voice classification is carried out on samples of two kinds of modes and decision fusion is carried out on a conditional probability to obtain a final emotion identification result. According to the invention, fusion of the voice emotion features and the facial expression features is carried out by using a decision fusion method, so that accurate data can be provided for corresponding conditional probability calculation carried out at the next step; and an emotion state of a detected object can be obtained precisely by using the method, so that accuracy and reliability of emotion identification can be improved.

The invention provides a method for training a model for generating natural languages for describing image contents, and a method for generating natural languages for describing image contents by adoption of the model. The training method comprises the following steps of: A1) taking global features and local features of images in an image training set as inputs of an attention mechanism, so as toobtain a fusion result comprising the global features and the local features at the same time; and A2) taking the fusion result and a word training set as inputs of a long short-term memory network, and training the attention mechanism and the long short-term memory network by utilizing a loss function so as to obtain a weight of the attention mechanism and a weight of the long short-term memory network, wherein the loss function is a function of a condition probability of the previous word or the ith word of a plurality of words in the content of a known image and a natural sentence for describing the image content, and i is equal to 1,to imax.

A method of determining a relation between search queries, includes: maintaining a database comprising a search session and a record about a search query which is received from a user terminal during the search session; recording and maintaining click rate information for each of the search queries in a predetermined storage unit; generating total search session number information by counting a total number of search sessions which is set during the time interval generating first search session number information by counting a number of search sessions where a first search query is received during the time interval, and generating second search session number information by counting a number of search sessions where a second search query is received during the time interval, by referring to the database; generating third search session number information by counting a number of search sessions where the first search query and the second search query are received during the time interval, by referring to the database; generating conditional probability information by using the first search session number information and the third search session number information; generating correlation information by using the total search session number information, the first search session number information, the second search session number information, and the third search session number information; querying click rate information of the second search query by referring to the storage unit; and determining a relation between the first search query and the second search query, based on the conditional probability information, the correlation information, and the click rate information.

The invention discloses a method for extracting information from an academic home page. The method comprises the following steps of: (1) finding an academic home page from Internet; (2) crawling and analyzing the academic home page, wherein the crawling of an irrelevant page is reduced by using a heuristic strategy so as to accelerate analysis speed; (3) analyzing the page into a form of documentobject module (DOM), and dividing according to attributes and contents of elements so as to acquire a cohesive text unit list; (4) identifying the text unit by using an information recognizer, wherein each information recognizer only identifies one information type, and performing subfield extraction on the text information; (5) performing association analysis on the extraction result, eliminating different meanings by using the association of the information, and complementing the missing field; and (6) matching the extraction result and a database, and eliminating the redundant data, wherein the extraction result is stored in a semantic database in a form of semantic data. In the method, by combination of heuristic rules, a machine learning method and a conditional probability model, academic information can be extracted efficiently and accurately from the academic home page.

Systems and Methods for multi-modal or multimedia image retrieval are provided. Automatic image annotation is achieved based on a probabilistic semantic model in which visual features and textual words are connected via a hidden layer comprising the semantic concepts to be discovered, to explicitly exploit the synergy between the two modalities. The association of visual features and textual words is determined in a Bayesian framework to provide confidence of the association. A hidden concept layer which connects the visual feature(s) and the words is discovered by fitting a generative model to the training image and annotation words. An Expectation-Maximization (EM) based iterative learning procedure determines the conditional probabilities of the visual features and the textual words given a hidden concept class. Based on the discovered hidden concept layer and the corresponding conditional probabilities, the image annotation and the text-to-image retrieval are performed using the Bayesian framework.

A method of underwriting group disability insurance for a policy period includes collecting medical claims data for the group to be underwritten, where each medical claim being related to a particular employee of the group. Morbidity categories are provided that categorize the medical claims in the medical claims data. A conditional probability model is developed and applied to the morbidity categories for each employee in the group using his medical claims, thereby calculating the expected conditional probability for each employee of incurring a disability during the policy period. A further statistical model of the estimated cost of the disability is developed and applied based on the employees' morbidity categories from the medical claims data. For each employee, an estimate of the expected cost of incurring a disability given their morbidity categories is derived from his medical claims data. Combining the expected conditional probability for each employee incurring a disability during the policy period with the estimate of the expected cost of that disability gives an estimate of the group's total disability exposure. Thereby, the expected disability exposure is used to determine a premium amount for disability insurance coverage during the policy period for the group.

A system and method for inferring informational goals and preferred level of details in answers in response to questions posed to computer-based information retrieval or question-answering systems is provided. The system includes a query subsystem that can receive an input query and extrinsic data associated with the query and which can output an answer to the query. The query subsystem accesses an inference model to retrieve conditional probabilities that certain informational goals are present. One application of the system includes determining a user's likely informational goals and then accessing a knowledge data store to retrieve responsive information. Determining a user's likely informational goals can include inferring a desired level of detail of answers to the query based on the application being employed by the user at the time the query is submitted. The system includes a natural language processor that parses queries into observable linguistic features and embedded semantic components that can be employed to retrieve the conditional probabilities from the inference model. The inference model is built by employing supervised learning and statistical analysis on a set of queries suitable to be presented to a question-answering system. Such a set of queries can be manipulated to produce different inference models based on demographic and/or localized linguistic data.

A data infrastructure for graph-based computing that combines the natural language expressiveness of the Semantic Web and the mathematical rigor of graph theory to discover meaningful associations across multiple sources towards computer-assisted serendipitous insight discovery. The process automatically integrates massive size datasets accessed using Semantic Web standards and technologies and normalizes data in graphs. The process generates a plurality of conditional probability distributions based on type-triple meta-data and triple statistics to model saliency and automatically construct and evaluate a plurality of sub-graphs based on the plurality of conditional probabilities for contextual-saliency. The process then renders a plurality of paths (i.e. sequence of associations) that model meaningful pairwise relations between objects of the normalized integrated data. The pluralities of conditional probabilities reveal and rank previously unknown associations between entities of user-interest in the knowledge graph.