US8818932
(Abstract)
"A method for creating（作成、作る）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（構築）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（一致、適合）grammatical constituents to semantic roles, and reorganizing data into clusters of entities with common attributes."
"A Bayesian network (BN), or directed graphical model, specifies a joint probability distribution（確率分布）over a collection of random variables（確率変数）as a graph encoding conditional independence relationship and a set of local distributions encoding probability information. Each node in the graph represents a random variable that is conditionally independent of its non-descendants given its parents. The local distributions at each node specify a set of probability distributions for the associated random variable, one for each combination of values for the node's parents. The local distributions implicitly encode a joint distribution over configurations of the random variables that satisfy the independence assumptions implied by the graph. In Bayesian learning, a prior distribution is defined over graph structures and local distributions, and the cases are used to infer a posterior distribution. The common approach is to assign a prior probability q(G) to each graph and independent Dirichlet distributions g(θ|G) for each of the local conditional distributions θ."
US8478629
"Referring back to FIG. 4, if a “performance measure” has violated a control limit, then the method proceeds to step 410. If a “risk event” has violated a control limit, then the method proceeds to step 420."
US9255988
"Referring again to FIG. 2, in block 35, a determination is made, by the processor, as to whether the cluster is stationary based on the classification identified in block 33. If the cluster is determined as stationary, the routine proceeds to block 36; otherwise, the routine proceeds to block 37."
US7981032
"In this case, the instruction is detected in step S205 and flow proceeds to step S206, in which the alarm is temporarily disabled for a specified period."
US7389347
"Referring now to FIG. 4, a flow diagram illustrates an active probing process, according to an embodiment of the present invention. More particularly, FIG. 4 describes（（図面が）説明する）an overall active probing process 400.
In step 402, the detection probes set and the localization probes set are selected. Next, in step 404, active diagnosis is started. Probes from detection probes set are run according to a schedule in order to detect problems (step 406). If a problem is detected (step 408), a problem localization process starts (410). Results of problem localization are reported and the process returns to step 406 to run scheduled detection probes set."
US7167587
"The program commences at step 202 and proceeds to step 204. In step 204, the process reads the stored training data for each class. The program then proceeds to step 206, where it waits for the input of an input pattern 18. Upon receipt of（受け取ると）an input pattern 18, the program proceeds to step 208, where a timer is set to govern the processing time used for the input pattern."
"The process then advances to step 220, where the computes a confidence value for its represented class best resembling the input pattern, reflecting the a posteriori probability that the selected class is the class associated with the input pattern."
US6336108
"The calculate mixed score process then determines if there are more variables in the selected set for processing (step 1010). If there are more variables to be processed, processing continues to step 1006."
