Abstract: In one embodiment, the present invention includes a method for identifying a deformable object of a scene of a computer game that is visible by an artificial intelligence (AI) character of the game, requesting a speculative physics simulation associated with the deformable object to determine a result of an action to the deformable object by the AI character, and selecting an action to be performed by the AI character, where the selection is based at least in part on the speculative physics simulation. Other embodiments are described and claimed.

Abstract: Feature values, which may be multi-dimensional, collected over successive time slices, are efficiently processed for use, for example, in known adaptive learning functions and event detection. A Markov chain in a recursive function to calculate imputed values for data points by use of a “nearest neighbor” matrix. Only data for the time slices currently required to perform computations must be stored. Earlier data need not be retained. A data selector, referred to herein for convenience as a window driver, selects successive cells of appropriate adjacent values in one or more dimensions to comprise an estimation set. The window driver effectively indexes tables of data to efficiently deliver input data to the matrix. In one form, feature inputs are divided into subgroups for parallel, pipelined processing.

Abstract: A method and system for solving an optimization problem comprising a plurality of dynamic constraints. A genetic algorithm is used to iteratively generate potential solutions to the problem. A constraint graph is used to model the plurality of dynamic constraints, and any potential solution that does not correspond to a connected subgraph of the constraint graph is infeasible and discarded. Real-time changes in dynamic constraints are incorporated by modification of the constraint graph between iterations of the genetic algorithm. An exemplary embodiment comprising the scheduling of air missions is presented.

Abstract: A device and a method for the analysis and troubleshooting of a system, based on the use of a model of the system, with application in particular in the area of industrial installations controlled by automatic control systems of the programmed or wired logic type. The method includes a stage for initialisation of the model, and a stage for the creation of a list of discordant variables whose value in the system differs from that predicted by the model. For each of the variables belonging to a discordance list, an initial list of suspect variables, suspected of having generated the discordant value, is created, and then a restricted list of suspect variables is obtained by filtration of the initial list.

Abstract: A system for determining that a document B is a candidate for near duplicate to a document A with a given similarity level th. The system includes a storage for providing two different functions on the documents, each function having a numeric function value. The system further includes a processor associated with the storage and configured to determine that the document B is a candidate for near duplicate to the document A, if a condition is met. The condition includes: for any function ƒi from among the two functions, ƒi(A)?ƒi(B)??i(ƒ,A,th).

Abstract: A statistical approach implementing Singular Value Decomposition (SVD) to a policy-based management system for autonomic and on-demand computing applications. The statistical approach empowers a class of applications that require policies to handle ambiguous conditions and allow the system to “evolve” in response to changing operation and environment conditions. In the system and method providing the statistical approach, observed event-policy associated data, which is represented by an event-policy matrix, is treated as a statistical problem with the assumption that there are some underlying or implicit higher order correlations among events and policies. The SVD approach enables such correlations to be modeled, extracted and modified. From these correlations, recommended policies can be selected or created without exact match of policy conditions.

Abstract: Classification of objects using the best boolean expression that represents the most optimal combination of the underlying features is disclosed.

Abstract: Methods and apparatus for an intelligent human-machine interface comprising an interface shell, system agents, function agents, a dynamic documentation system, and a layering architecture. The system agents include dynamic, knowledge-based software object sub-agents that model and track the state of a work area. The function agents model, track, and facilitate work area functions. The interface shell provides a hardware and software interface between the system agents and the function agents. The layering architecture comprises a tracking layer, an equipment and supply management layer, a coordination layer, and a situational awareness layer, and an oversight layer. The oversight layer combines information from the situational awareness layer with process rule sets contained in the function agents to determine if processes are being performed correctly.

Abstract: A method for training classifiers for Computer-Aided Detection in medical images includes providing an image feature training set {(xi, yi)}i=1l, wherein xi?Rd are input feature variables and yi?{?1,1} are class labels, and a cascade of K classifiers to be trained, minimizing, for each classifier k, a first cost function to initialize an ?k0 associated with each classifier k, fixing all classifiers except classifier k and minimizing a second cost function to solve for ?kc for a counter value c using the training dataset {(xik, yi)}i=1l, calculating a third cost function Jc(?lc, . . . , ?Kc) for each classifier k, and comparing Jc with a previous iteration Jc?1, wherein if Jc?Jc?1 is less than a predetermined tolerance, said classifier training is completed.

Abstract: An example embodiment provides a process relating to an inference engine for enterprise systems and enterprise contracts. In the example, the inference engine receives a fact and stores it in an input queue. Then the inference engine retrieves the fact from the input queue and routes the fact to a schedule queue on the basis of the fact's processing schedule and a condition that is part of a rule. The inference engine retrieves the fact from the schedule queue in accordance with the processing schedule and routes the fact to a rule executer on the basis of the fact's contents. The rule executer applies an action to the fact, where the action is also part of the rule and the action transforms the fact or creates new facts. Then the inference engine routes the transformed fact or new facts to the input queue and possibly to a persistent storage device.

Abstract: A system and method that optimizes reduce operations by consolidating the operation into a limited number of participating processes and then distributing the results back to all processes to optimize large message global reduce operations on non power-of-two processes. The method divides a group of processes into subgroups, performs paired exchange and local reduce operations at some of the processes to obtain half vectors of partial reduce results, consolidates partial reduce results into a set of regaining processes, performs successive recursive halving and recursive doubling at a set of remaining processes until each process in the set of remaining process has a half vector of the complete result, and provides a full complete result at every process.

Abstract: A method and system for determining whether to alter a firewall configuration. Message flow data associated with a message packet blocked by a firewall is received. The packet was blocked based on the firewall not having a message flow rule that permitted passage of the message packet. Risk values associated with a source network, destination network and destination port are identified by the message flow data. Based on the risk values, an electronic recommendation indicating whether to add to the firewall a message flow rule that permits the message flow to pass is determined and generated.

Abstract: The invention is a technique for performing sampling in connection with Markov Chain Monte Carlo simulations in which no attempt is made to limit the selected samples to a selected slice of the entire sample domain, as is typical in Markov Chain Monte Carlo sampling. Rather, samples are taken from the entire domain and any samples that fall below a randomly selected probability density level are discarded.

Abstract: A classification system is described for resiliently classifying data. In various embodiments, the classification system constructs a combined classifier based on multiple classifiers that are constructed to classify a set of training data. The combined classifier can be constructed in parallel with the multiple classifiers and applied to classify data.

Abstract: A system and method for automatic task prioritization comprises one or more task queues, a monitoring module, and a decision engine. The decision engine receives tasks and assigns a priority code to each task. Each task is inserted into one of the task queues according to its priority code and the priority codes of other tasks which may be present in the task queues. Agents select tasks to perform from the task queue according to priority guidelines established by a system user. The monitoring module may monitor the order of the tasks selected by the agents and each task's priority code. The monitoring module feeds this information back to the decision module. In one embodiment, the decision module is a learning system that updates stored data using feedback from the monitoring module to determine priority of each task.

Abstract: Provided are a method, system, and article of manufacture for using a data mining algorithm to discover data rules. A data set including multiple records is processed to generate data rules for the data set. Each record has a record format including a plurality of fields and each rule provides a predicted condition for one field based on at least one predictor condition in at least one other field. The generated data rules are provided to a user interface to enable a user to edit the generated data rules. The data rules are stored in a rule repository to be available to use to validate data sets having the record format.

Abstract: In accordance with the embodiments described herein, a method and system for adaptive knowledge lifecycle management, and methods for delivering adaptive knowledge lifecycle management services to customers, is disclosed. Adaptive knowledge lifecycle management provides a means for beneficially adapting knowledge assets and their collective structures over time. It also provides a means for generating an estimate of the value of knowledge assets and their associated organizing topologies, as well as providing a means to determine changes in value corresponding to changes to the knowledge assets and their associated organizing topologies.

Abstract: A method for providing a polyhierarchical classification includes identifying properties of objects useful for distinguishing objects under classification. A plurality of criteria are identified for specializing the identified properties. A form is chosen for attributive expressions that describe classification categories. The attributive expressions are customizable and encode compositions of object properties in terms of attributes from the plurality of criteria. A domain of applicability is identified for each criterion that is representable by attributive expressions, and a dependence relationship between criteria is defined by the inclusion of attributes in the attributive expressions, where a selected criterion depends on another criterion if its domain of applicability includes at least one attribute by the other criterion. A generating polyhierarchy of criteria is automatically established by the dependence relationships between the criteria.

Abstract: A method and system to minimize to redundancy in testing a new software system by utilizing a knowledge base is provided. The knowledge base may be represented in a form of a decision tree including leaf nodes which store previously tested optimal configurations. When the knowledge base does not contain an optimal configuration for the new software, an optimal solution for the new software system may be determined by adding a further decision tree branch to the initial decision tree. A desirable tree branch may be selected from a table including a predetermined set of desirable tree branches for each category of software. As a result, the number of solutions to be tested on the input software may be limited and cost and effort to run redundant tests on a software system may be reduced.

Abstract: A set of objects having facts is established. Facts of objects having positions in an order are identified. Some facts explicitly describe the positions in the linear order, while are facts do not explicitly describe the positions. The facts are presented in the order on a linear graph, such as a timeline. Facts of the objects describing geographic positions are presented on a map.