30 results about "Causal link" patented technology

A method for creating a predictive model is disclosed herein, including the steps of determining trends and patterns in electronic data, using at least a first machine language algorithm, refining the determination of the algorithm, searching for social models that describe the identified trends and patterns using at least a second machine language algorithm, verifying causal links, constructing at least one model about human node behavior and interactions, utilizing the social models to do at least one of the following: validate hypotheses, predict future behavior, and examine hypothetical scenarios, and automatically updating predictions when new data is introduced.

The present invention relates to control, monitoring, and automation. The present invention more specifically relates to pattern-based intelligent control, monitoring and automation. The invention performs pattern-based monitoring. It collects signal data from one or more signals. The signal data define signal data streams. It then transforms each of the signal data streams into trends. It also discovers patterns based on the trends within each signal data stream and / or across the signal data streams. The patterns are optionally used for diagnostics and root cause analysis, online plant monitoring and operation control, plant optimization, and other environments where a causal link or correlation may exist between related inputs, states and / or outputs.

A computer-readable medium to direct a computer to function in a specified manner includes executable instructions to construct a strategy map characterizing a presumed causal link between a first business metric and a second business metric. A first pattern within data associated with the first business metric is identified; a second pattern within data associated with the second business metric is located. Any relationship between the first pattern and the second pattern is specified.

A method of adapting a security configuration of a data processing application at runtime, and a system, together with its computing architecture, are disclosed. The system stores a causal network comprising a plurality of nodes and a plurality of incoming and outgoing causal links associated therewith, wherein each node of the causal network is associated with a security concern or a requirement that can be affected by any configuration of the security controls. The current value of assets nodes, as well as those of the security concerns that can be affected by monitored contextual factors, are updated. The control nodes corresponding to the security controls is updated according to the security configuration whose utility is evaluated by the causal network. The node corresponding to the at least one variable is updated with the determined current value, which is propagated through the causal network through the causal links associated with the updated node. The security configuration with the highest utility is selected and replaces the actual configuration by activating and / or deactivating the security functions corresponding to security control nodes enabled / disabled in the selected security configuration.

According to one embodiment, a method comprises evaluating messages between nodes of a distributed computing environment. Based on timing relationships of the messages, probability of causal links between pairs of messages determined. Based on the determined probability, at least one causal path comprising a plurality of causal links is inferred. Further, an overall probability that the inferred causal path is accurate is determined. In certain embodiments, the overall probability is output for the corresponding causal path to provide an indication of the confidence of the accuracy of such inferred causal path.

A method for creating a predictive model is disclosed herein, including the steps of determining trends and patterns in electronic data, using at least a first machine language algorithm, refining the determination of the algorithm, searching for social models that describe the identified trends and patterns using at least a second machine language algorithm, verifying causal links, constructing at least one model about human node behavior and interactions, utilizing the social models to do at least one of the following: validate hypotheses, predict future behavior, and examine hypothetical scenarios, automatically updating predictions when new data is introduced, using probabilistic techniques to learn hierarchical structure in unstructured text, continuously updating a set of themes, examining grammatical rules of each component of text, matching grammatical constituents to semantic roles, and reorganizing data into clusters of entities with common attributes.

The invention discloses an aviation safety accident causal relationship extraction method. The method comprises the steps of text data preprocessing; extracting explicit causal relationships; Extracting an implicit causal relationship; and generating a causal relationship chain. The method has the advantages that the mode matching method is applied to extraction of the explicit causal relationshipof the aviation safety accident, and the bidirectional LSTM method based on the self-attention mechanism is adopted; applying the At-BiLstm + PI to implicit causal relationship extraction, giving a single accident causal relationship chain on the basis that identification and extraction of a single aviation safety accident causal relationship pair are effectively achieved, and good data and method support is provided for scene reproduction and comprehensive analysis of aviation safety accidents.

Summarization of an input story can be accomplished through identification of causal relationships, both explicit and implicit. The input story is transformed into an interpretation sequence, using syntactic cues, and common sense knowledge of an average reader. The interpretation sequence is a time ordered semantic representation of the input story, comprised of story elements. The common sense knowledge includes inference rules, which, from story elements already present, can add additional story elements to the interpretation sequence. Application of inference rules, based on type, can be prioritized. Summarization of the interpretation sequence can be accomplished by the selection of explicit story elements, according to a connection-based strategy, or a concept-based strategy. Regarding a concept-based strategy, metrics can be applied, to select the concepts for contra-causal searching of the interpretation sequence. Options can be provided, for the exclusion of means, or the inclusion of implicit, story elements in the output summary.

A utility that can be embedded within a system or application, or run as a stand-alone utility, provides a method for identifying causal relationships for one or more events. For instance, if an anomaly is occurring for a subset of events in a series of events, certain information about potential causes can be collected in conjunction with the series of events and, the information can be analyzed to determine if there is a cause and affect relationship leading to the occurrence of the anomalies.

The invention discloses a safety state confirmation method based on construction parameters. Different construction risk factors are selected for different construction nodes in a project to serve as random variables for safety state confirmation; determining a sequence and a relationship of the random variables in different construction nodes, obtaining the sequence and the relationship of the random variables by training a convolutional neural network and / or determining the sequence and the relationship of the variables according to a causal relationship, counting occurrence frequencies of various accidents at different construction nodes, and constructing a three-layer risk accident causal chain, distributing conditional probability distribution of different nodes in the causal chain according to the obtained risk accident causal chain, sorting possible risk accidents and occurrence probabilities according to accident severity to obtain first sorting data and second sorting data, and when a constructor enters a construction node at the position, determining the occurrence probability of the risk accidents according to the first sorting data and the second sorting data. And confirming the safety state of the personnel according to the sequence of the first sorted data and the second sorted data.

The invention relates to a causal relationship test and micro-service index prediction alarm method. The method comprises service index causal relationship discovery based on Granger causal relationship test and multi-index prediction based on Attention LSTM. According to the method, the indexes having the causal relationship with the to-be-predicted index are found through the Granger causal relationship test and jointly participate in prediction, so that the prediction accuracy is improved. And meanwhile, aiming at the problems that the index sequence in the micro-service scene is relatively long, the causal relationship exists locally among the indexes, and the overall causal relationship is not strong, the Granger causal relationship test is improved, and the causal relationship is calculated in a segmented increment manner. In the practical application of the method, when a new value is added into the time sequence, only the causal relationship of the increment needs to be calculated, and the historical data does not need to be re-calculated, so that the calculation amount is reduced, and the causal relationship discovery efficiency in the micro-service scene is improved.

The invention discloses a road transportation safety risk real-time evaluation method based on data mining, and the method is calculated based on the following model: R = P.M, R is a safety risk evaluation value, P is a risk source or accident occurrence probability evaluation index, and M is a risk source or accident possible damage consequence severity evaluation index. According to the invention, based on mining of a large amount of text data, dynamic real-time acquisition of P and M is realized, real-time evaluation of risk values is obtained, real-time dynamic change of risk evaluation along with risk source coupling is realized, risk evaluation considering risk source causal relationships is realized, and automatic updating of evaluation thresholds is realized.

The invention discloses a causal link analysis method, causal link analysis equipment and a computer readable storage medium. The method comprises the steps: carrying out the recording and analysis ofa production process of a specified object, and obtaining a plurality of specified variables corresponding to the production process; determining an undirected correlation link among the plurality ofspecified variables based on a relevancy judgment rule; predicting the undirected correlation link according to a causal probability prediction model to obtain a directed causal probability value corresponding to the undirected correlation link; determining a directed causal link among the plurality of specified variables according to the directed causal probability values, wherein the directed causal link is used for representing a causal relationship of the specified object production process; and the mutual influence sequence of the specified variables in the production process can be known.

Embodiments of the present invention provide a failure prediction method and device for a cluster system. The method includes: performing cluster analysis on event sequence pattern sets according to failure event sets of the cluster system to obtain event sequence clusters; performing causal correlation analysis on event sequence clusters to obtain causal correlation path clusters; performing failure analysis on causal correlation path clusters to obtain The failure mode of the cluster system; according to the failure mode, the failure prediction of the cluster system is carried out. The failure prediction method and device of the cluster system provided by the embodiments of the present invention cluster the event sequence pattern set to obtain the causal correlation path cluster, thereby obtaining the failure mode of the cluster system, and predicting the failure of the cluster system according to the obtained failure mode. Through in-depth analysis and mining of cluster system logs, reliable failure rules for failure prediction can be obtained, thereby improving the recall rate of failure prediction.

The invention relates to a product quality characteristic error traceability analysis method based on multi-source perception, which comprises the following steps of establishing causal relationship between processes by using historical data; screening the historical data so as to compose the screened historical data into a sample space; taking the sample space as the analysis standard, and monitoring the real-time data by a T2 control chart method; performing the orthogonal decomposition on the outbound T2 value according to the causality to obtain decomposition terms; carrying out the out-of-bounds analysis on the T2 statistic value of the decomposition term, and locating the problem procedure. The method of the invention utilizes a large amount of multiple historical data to complete the analysis of the correlation relation between the processes, and reduces the influence brought by the subjective factors. The out-of-bounds T2 value is decomposed according to the relationship directed graph obtained from the analysis process of the relationship between processes, so as to determine the problem process, achieve the accurate grasp of the process, and then achieve the purpose of problem traceability.

A system for identifying a causal link, the system including a diagnostic generator module configured to receive a first user symptom datum, receive diagnostic training data, and generate using a supervised machine-learning process a diagnostic model that outputs a first prognosis. The system includes a prognostic chaining module configured to receive an expert input dataset, receive the first user symptom datum and the first prognosis, generate a gaussian mixture clustering model and identify a first causal link chained to the first prognosis. The system includes a causal link module configured to receive the first prognosis chained to the first causal link, receive a second prognosis chained to a second causal link, and evaluate the first causal link and the second causal link to calculate a degree of similarity between the first causal link and the second causal link.

A system for identifying a causal link, the system including a diagnostic generator module configured to receive a first user symptom datum, receive diagnostic training data, and generate using a supervised machine-learning process a diagnostic model that outputs a first prognosis. The system includes a prognostic chaining module configured to receive an expert input dataset, receive the first user symptom datum and the first prognosis, generate a gaussian mixture clustering model and identify a first causal link chained to the first prognosis. The system includes a causal link module configured to receive the first prognosis chained to the first causal link, receive a second prognosis chained to a second causal link, and evaluate the first causal link and the second causal link to calculate a degree of similarity between the first causal link and the second causal link.

Type 1 diabetes mellitus is characterized by loss of pancreatic insulin-producing beta cells, resulting in insulin deficiency. The usual cause of this beta cell loss is autoimmune destruction. The inventors provide the first evidence of a causal link between influenza virus infection and the development of type 1 diabetes and / or pancreatitis. This causal link between infection and type 1 diabetes and / or pancreatitis provides various therapeutic, prophylactic and diagnostic opportunities.

The invention relates to a method and a system for extracting causal relationships between affairs. The method comprises the following steps: segmenting a to-be-processed text through punctuations, and obtaining at least one clause; according to the clauses and predefined causal links, judging whether causal relationships exist among the clauses or not, generating a first causal set, and defining the corresponding clauses as cause fragments and result fragments respectively; according to the cause fragments and result fragments and by using dependency grammar, dividing the cause fragments and result fragments into at least one candidate event, judging causal relationships between candidate events, generating a second causal set, and obtaining sub-cause fragments and sub-result fragments corresponding to the cause fragments and result fragments; extracting sub-result fragments in the cause fragments as causes of a text core relationship, and extracting sub-cause fragments in the result fragments as results in the text core relationship. According to the method and the system, a causal relationship of a subsequent cause and a preceding result can be identified, and the cause and the result are accurately extracted, so that the effectiveness of the cause and effect relationship identification is improved.

The invention discloses a dynamic medical data oriented causal feature extraction method, which comprises the following steps: 1, extracting features of a medical data set, and constructing / updating a feature set of the medical data; 2, according to the extracted features and the current medical data, summarizing the medical data information in the form of statistical information by constructing / updating a full-dimensional tree structure; 3, assigning an initial feature subset of the causal feature extraction by using an existing key feature (factor) set, namely initializing an optimal subset; 4, selecting an optimal feature subset by utilizing a causal relationship inference theory and algorithm through iteration of addition and deletion operations, namely obtaining key feature information of the current medical data; and 5, when new medical data arrives, repeating the steps 1-4, and selecting real-time key features from the dynamic medical data. According to the method, the key features can be found from the dynamic medical data, and the extraction of the key features is more interpretable and robust.

The invention discloses a medical field causal relationship extraction method fusing radical information, and relates to the technical field of data mining, and the method comprises the following steps: obtaining a Chinese medical field text data set through a web crawler, preprocessing the obtained data, converting English professional nouns in a text into Chinese by adopting a Google translation technology, and extracting the Chinese medical field text data set; the method comprises the following steps: acquiring radicals of all characters by using an online Xinhua dictionary, performing incremental training on the radicals by using a Word2Vec architecture to obtain a radical feature representation, and splicing a radical feature vector and a character feature vector to obtain character features fused with radical information.

A massive clonal expansion of activated CD8+ T-cells with increased frequency of HPV 16-specific CD8+ T-cells was discovered to be a characteristic of oral lichen planus (OLP), indicating a causal link between HPV infection and the dysimmune process. The invention relates to compositions and methods for the diagnosis and treatment of OLP patients.

Systems and methods are described herein for logging system events within an electronic machine using an event log structured as a collection of tree-like cause and effect graphs. An event to be logged may be received. A new event node may be created within the event log for the received event. One or more existing event nodes within the event log may be identified as having possibly caused the received event. One or more causal links may be created within the event log between the new event node and the one or more identified existing event nodes. The new event node may be stored as an unattached root node in response to not identifying an existing event node that may have caused the received event.