Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training a neural network used to select actions performed by a reinforcement learning agent interacting with an environment. In one aspect, a method includes maintaining a replay memory, where the replay memory stores pieces of experience data generated as a result of the reinforcement learning agent interacting with the environment. Each piece of experience data is associated with a respective expected learning progress measure that is a measure of an expected amount of progress made in the training of the neural network if the neural network is trained on the piece of experience data. The method further includes selecting a piece of experience data from the replay memory by prioritizing for selection pieces of experience data having relatively higher expected learning progress measures and training the neural network on the selected piece of experience data.

Abstract: In accordance with the invention, a system, method, and apparatus for analyzing advertisement-related data are presented, which may include receiving data related to an aspect of an advertisement and modeling the aspect of the advertisement with a mathematical model. The mathematical model may include a control-signal-related component, a control-signal-independent component, and an error component. Each component may be updated based on at least one of a control signal, the received data, and a previous state of at least one of the components. An updated model may be created base on the updated components. The system, method, and apparatus may also include predicting the aspect of the advertisement using the updated model. Exemplary aspects of and data related to the advertisement may include one or more of the following: a number of impressions, “clicks,” or “conversions” and/or the impression-to-conversion, impression-to-click, or click-to-conversion ratios.

Abstract: Methods, apparatuses and articles of manufacture for use in a mobile device to determine whether a dynamic user profile is to transition from a first state to a second state based, at least in part, on one or more sensed indicators. The dynamic user profile may be indicative of one or more current inferable user behavior contexts for a user co-located with the mobile device. The mobile device may transition a dynamic user profile from a first state to a second state, in response to a determination that the dynamic user profile is to transition from the first state to the second state, and operatively affect one or more functions performed, at least in part, by the mobile device based, at least in part, on the transition of the dynamic user profile to the second state.

Abstract: In accordance with the invention, a system, method, and apparatus for analyzing advertisement-related data are presented, which may include receiving data related to an aspect of an advertisement and modeling the aspect of the advertisement with a mathematical model. The mathematical model may include a control-signal-related component, a control-signal-independent component, and an error component. Each component may be updated based on at least one of a control signal, the received data, and a previous state of at least one of the components. An updated model may be created based on the updated components. The system, method, and apparatus may also include predicting the aspect of the advertisement using the updated model. Exemplary aspects of and data related to the advertisement may include one or more of the following: a number of impressions, “clicks,” or “conversions” and/or the impression-to-conversion, impression-to-click, or click-to-conversion ratios.

Abstract: The method of predicting gas composition in a multistage separator includes solutions to the regression problem of gas composition prediction that are developed using an ensemble of hybrid computational intelligence (CI) models. Three separate homogeneous and one heterogeneous ensemble of hybrid computational intelligence (EHCI) models are developed using a parallel scheme. The homogeneous models have the same types of CI models used as base learners, and the heterogeneous model has of different types of CI models used as base learners. Various popular CI models, including multi-layer perceptron (MLP), support vector regression (SVR) and adaptive neuro-fuzzy inference system (ANFIS), are used as base learners of ensemble models.

Abstract: An industrial controller simulation system is provided. The system includes a simulation component that enables modeling of an industrial controller system. A suggestion component offers automated recommendations in accordance with the modeling of the industrial controller system.

Abstract: A system and method provides a solution to the problem of applying end-to-end requirements of connectivity, security, reliability and performance to configure a network and ultimately assign network components to the network. All requirements are modeled as constraints and a constraint solver does the resolution. Not every constraint to be solved is solved by the model-finder. Instead, we “factor away” subsets of a constraint that can be efficiently solved via a special-purpose constraint solver, such as an SQL/Prolog engine, linear programming system, or even an algorithm, leaving behind a constraint that truly requires the power of model-finding, and that is often efficiently solvable by existing model-finders. Such constraints are compiled into quantifier-free constraints that are Boolean combinations of constraints of two forms x=y and x=c where x, y are variables and c is a constant. Such constraints can be efficiently solved by modern SAT-based model-finders.

Abstract: The method for generating an integrated guidance law for aerodynamic missiles uses a strength Pareto evolutionary algorithm (SPEA)-based approach for generating an integrated fuzzy guidance law, which includes three separate fuzzy controllers. Each of these fuzzy controllers is activated in a unique region of missile interception. The distribution of membership functions and the associated rules are obtained by solving a nonlinear constrained multi-objective optimization problem in which final time, energy consumption, and miss distance are treated as competing objectives. A Tabu search is utilized to build a library of initial feasible solutions for the multi-objective optimization algorithm. Additionally, a hierarchical clustering technique is utilized to provide the decision maker with a representative and manageable Pareto-optimal set without destroying the characteristics of the trade-off front. A fuzzy-based system is employed to extract the best compromise solution over the trade-off curve.

Abstract: The control optimization method for helicopters carrying suspended loads during hover flight utilizes a controller based on time-delayed feedback of the load swing angles. The controller outputs include additional displacements, which are added to the helicopter trajectory in the longitudinal and lateral directions. This simple implementation requires only a small modification to the software of the helicopter position controller. Moreover, the implementation of this controller does not need rates of the swing angles. The parameters of the controllers are optimized using the method of particle swarms by minimizing an index that is a function of the history of the load swing. Simulation results show the effectiveness of the controller in suppressing the swing of the slung load while stabilizing the helicopter.

Abstract: The fuzzy logic-based control method for helicopters carrying suspended loads utilizes a controller based on fuzzy logic membership distributions of sets of load swing angles. The anti-swing controller is fuzzy-based and has controller outputs that include additional displacements added to the helicopter trajectory in the longitudinal and lateral directions. This simple implementation requires only a small modification to the software of the helicopter position controller. The membership functions govern control parameters that are optimized using a particle swarm algorithm. The rules of the anti-swing controller are derived to mimic the performance of a time-delayed feedback controller. A tracking controller stabilizes the helicopter and tracks the trajectory generated by the anti-swing controller.

Abstract: A user-concerned information recommendation system and method considering user's watching or listening time and the maximum playing time of contents are disclosed. The user-concerned information provision system includes a plurality of user terminals to provide contents transmitted from an external server to a user, a user-concerned information inference server to infer an association relationship between the contents based on information of maximum playing time and actual playing time of the contents provided to the user terminals, and a content provision server to provide a content requested by an arbitrary one of the user terminals and other contents associated with the requested content according to the inferred association relationship to the arbitrary one of the user terminals when receiving a request of the content from the arbitrary one of the user terminals, thereby providing more accurate user-concerned information to the user.

Abstract: Documents to be accessed over a computer network are stored in network data storage such that each document is associated with a document archive and a task is associated with each stored document. An escalation list is created for each task and is associated with the document archive so as to specify an ordered list of escalation items to be satisfied until fulfillment of the task. When a parameter associated with each escalation list item has been satisfied, process operations associated with the next escalation list item are performed. Fulfillment of the task results in appending a verification of task fulfillment to the document archive.

Abstract: The present invention relates to a system and a method for automatically determining the relationship between a plurality of tasks comprising a process and then detecting any abnormalities in the execution of the process.

Abstract: A method and an apparatus for predicting intake manifold pressure are presented, to compensate for a large lag or a large time delay without producing an overshot or discontinuous behaviors of a predicted value. The method comprises the step of obtaining a difference of values of a variable to be predicted and a difference of values of another variable ahead of the variable to be predicted. The method further comprises the step of filtering the differences with adaptive filters. The method further comprises the step of obtaining a predicted difference of values of the variable to be predicted, through algorithm of estimation with fuzzy reasoning. The method further comprises the step of adding the predicted difference of values of the variable to be predicted, to a current value of the variable to be predicted, to obtain a predicted value of the variable to be predicted.

Abstract: A control procedure is provided for use during a regulation stage and according to a set point of a physical dynamic system, the set point being subject to, whilst operating, the influence of several physical quantities represented by input parameters, and adopting a behavior defined by at least a first physical output parameter, obliged to take a value represented by the set point, the first output parameter being linked to at least a first of the input parameters by a first transfer function of the system. According to the control procedure, a characterization stage is implemented in which at least a first inverse transfer function linking the first input parameter to the first output parameter is experimentally determined. A modeling stage is implemented in which the first inverse transfer function is translated through a fuzzy logic model in the form of a first set of ranges of the first output parameter, to each of which is attributed a specific value of the first input parameter.

Abstract: A system is controlled by an actor-critic based fuzzy reinforcement learning algorithm that provides instructions to a processor of the system for applying actor-critic based fuzzy reinforcement learning. The system includes a database of fuzzy-logic rules for mapping input data to output commands for modifying a system state, and a reinforcement learning algorithm for updating the fuzzy-logic rules database based on effects on the system state of the output commands mapped from the input data. The reinforcement learning algorithm is configured to converge at least one parameter of the system state to at least approximately an optimum value following multiple mapping and updating iterations. The reinforcement learning algorithm may be based on an update equation including a derivative with respect to at least one parameter of a logarithm of a probability function for taking a selected action when a selected state is encountered.

Abstract: The invention is a method for augmenting images to highlight regions where objects of interest are likely to appear. The method uses fuzzy set theory to classify locations that may contain objects of interest, uses fog volumes to represent the fuzzy sets in three dimensions, and renders these fog volumes from the vantage point of the image to provide a colored overlay with which to augment the image. Where fog volumes of different colors overlap in an image, the invention renders the overlapping area using cross-hatching, so that the multiple colors appear side by side rather than blending to make a third color.

Abstract: Machine operators rely on intuition and experience to evaluate vehicle performance. As we increasingly turn to automation, it is important to automatically evaluate sensor data and system performance. Fuzzy logic allows us to take advantage of domain knowledge to evaluate data and to describe a system linguistically. In this application a fuzzy quality module evaluates output from a noisy sensor. It compares the sensor data with an estimated sensor value, and based on that comparison, dynamically adapts a fuzzy logic rulebase used to determine whether the sensor value is valid or not.

Abstract: A computer implemented method and apparatus for simulating an intelligent tutor for interactive adaptive training of learners in any domain includes a Domain Module, a Tutor Module and an Interface. Items to be learned, and their prerequisite and other dependency relationships are represented in a fuzzy graph, together with a fuzzy logic computational engine, which dynamically adapts the available sequence of training actions (such as presentations/explanations, simulations, exercises and tasks/questions) to a current assessment of the learner's knowledge skill, the level of difficulty of the presented material, and preferences and learning style of the individual learner. Fuzzy logic is used as the basis of arc weightings, and the computations, but the general methodology is applicable to other approaches to weighting in computation.

Abstract: A system and method is provided for monitoring the operating condition of a pump by evaluating fault data encoded in the instantaneous current of the motor driving the pump. The data is converted to a frequency spectrum which is analyzed to create a fault signature having fault attributes relating to various fault conditions associated with the pump. The fault signature is then input to a neural network that operates in conjunction with a preprocessing and post processing module to perform decisions and output those decisions to a user interface. A stand alone module is also provided that includes an adaptive preprocessing module, a one-shot unsupervised neural network and a fuzzy based expert system to provide a decision making module that operates with limited human supervision.

Abstract: An apparatus and method of generating an application for a control system. A control process is defined by a physical model and a topological model. An application generator utilizes the physical and topological models to generate an application for the control system.

Abstract: This invention proposes a novel nonlinear filter called adaptive weighted fuzzy mean (AWFM) filter which is capable of removing high density Gaussian impulse noises in polluted images. Via a single epoch, weights of the filter are departed according to a set of fuzzy rules, which are constructed by a proposed procedure. All the fuzzy rules in the knowledge base infer concurrently for input nature estimation and filter adaptation. Statistic analyses ensure that this AWFM filtering possesses robust and consistent performance even when images are thoroughly polluted. The AWFM filter is synthesized with generic LR fuzzy cells which adopts CMOS analog current mode technology to subject high speed, low power, and small circuit area objectives. Simulations show that parallel inferences realized by the generic LR fuzzy cells, each of which performs 20 M FLIPS (fuzzy logic inferences per second) promise that the synthesized AWFM filter is able to process up to 50 256×256 images per second.

Abstract: Method and apparatus for combining a plurality of overlapping policy-based controllers. System also applicable to policy-based process servers. System combines controllers by combining the respective policy information. System combines a plurality of policy-based sub-controllers by combining the associated distributional information contained in the associated sub-policies. An iterative mixture mechanism with temporal persistence regulates the relative contribution of the sub-policies smoothly over time thereby allowing smooth transition of control from one control regime to another. The system provides for modular detection and resolution of conflicts that may arise as a result of combining otherwise incompatible sub-policies. Preferred embodiment performs mixture method in policy space. Another embodiment applies mixture method to value functions associated with each sub-server.

Abstract: An advanced control block that implements multiple-input/multiple-output control, such as model predictive control, within a process control system is initiated by creating an initial control block having generic control logic and desired control inputs and control outputs communicatively connected to process outputs and process inputs within a process control routine. A waveform generator within the control block systematically upsets each of the process inputs via the control block outputs using excitation waveforms designed for use in developing a process model. At the same time, a data collection routine collects data indicating the response of each of the process outputs to the waveforms delivered at each of the process inputs. After sufficient data has been collected, a process modeling routine generates a process model from the collected data and a control logic parameter creation routine creates control logic parameters for the control logic from the process model.

Abstract: A fuzzy processor with an improved architecture. The fuzzy processor includes a fuzzy rule processor, an internal fuzzy instruction memory, an internal knowledge base memory, an arithmetic-logic unit, a control unit that can execute non-fuzzy instructions that are typical of conventional microprocessors, and an internal memory for storing the non-fuzzy instructions. The improved fuzzy processor architecture has an ability to load other knowledge bases and other fuzzy rules from outside the processor concurrently and transparently with respect to instruction processing. The processor can also process both fuzzy instructions and non-fuzzy instructions, can perform conditional and unconditional jumps within a set of fuzzy rules that are being processed, and can conditionally swap the knowledge base or the set of rules that are to be processed.

Abstract: An optimized method of computing the value of the degree of membership of a fuzzy variable defined within a universe of discourse that is discreted into a finite number of points by way of a membership function thereof, wherein the membership function is quantified into a finite number of levels corresponding to a finite number of degrees of truth, and is stored as a characteristic value of each subset of fuzzy variable values being all mirrored in one value of said degree of membership corresponding to one of said levels. The computing method includes generating a binary sequence; generating an address signal from the bits in the binary sequence; reading the contents of the memory storing the membership functions at each address signal to obtain a characteristic value; and comparing the characteristic value with the value of a fuzzy input variable.

Abstract: A method for transformation of fuzzy logic (FS) into a neural network (NN), in which, in order to form a defuzzified output value (y2) from normalized single-element functions (F1 . . . Fm), the single-element functions (F1 . . . Fm) are each assigned a singleton position (A1 . . . Am) and at least one singleton weighting factor (R1 . . . Rn), those singleton weighting factors (R1 . . . Rn) which are assigned to the same single-element function (F1 . . . Fm) are additively linked, and the singleton weighting factors (R1 . . . Rn) and the additively linked singleton weighting factors (R1 . . . Rn) are weighted via the corresponding singleton positions (A1 . . . Am) and are additively linked in order to form the defuzzified output value (y2). One advantage of the method according to the invention is that the singleton positions (A1 . . .

Abstract: A fuzzy processor which can be programmed and expanded is disclosed. The fuzzy processor has a membership function I/O circuit for inputting and outputting a plurality of membership functions respectively corresponding to each one of a plurality of features of each one of a plurality of standard patterns. A feature decoder receives a to-be-recognized pattern having a plurality of input features for generating a plurality of feature values. A membership function generator stores the plurality of membership functions and receives the plurality of feature values to generate a plurality of current-type membership degrees for the plurality of input features corresponding to the plurality of standard patterns respectively. A plurality of accumulators receive the plurality of current-type membership degrees respectively for generating a plurality of synthesis membership degrees.

Abstract: An apparatus for determining the state of a dynamic system is disclosed. The apparatus comprises one or more sensors operable to sample physical parameters of a system, and a mode interpreter. The mode interpreter is operable to receive data from the sensor and determine the state of the dynamic system using hypertrapezoidal membership functions.

Abstract: The fuzzy inference circuit includes a sensor sensing a state of a control object, and a charge coupled device storing fuzzy rule function values for at least one fuzzy rule function. The charge coupled device further performs a fuzzy inference on output of the sensor using the fuzzy rule function values to generate a control signal for controlling the control object.

Abstract: A current-type fuzzy processor is disclosed. The fuzzy processor comprises a characteristic decoder for decoding a to-be-identified pattern, a plurality of membership function generators for generating a plurality of membership degrees for the to-be-identified pattern, a weighting decoder for generating a weighting value, a plurality of weighting adjustment circuits for performing weighting operations on the plurality of membership degrees according to the weighting value to generate a plurality of weighted membership degrees, a plurality of accumulators for performing accumulation operations on the plurality of weighted membership degrees to generate a plurality of synthesis membership degrees, a WTA circuit for outputting the plurality of synthesis membership degrees in an order of magnitude; and a timing and control logic for generating timing and control signals for the fuzzy processor.

Abstract: An apparatus over or under approximates the result of dividing a binary number representing an integer 2n. The division by 2n is performed by truncating the n least significant bits of the integer. In order to over or under approximate the result, the nth truncated bit, i.e., the most significant bit of the n-truncated less significant bits, is added to the integer represented by the remaining non-truncated bits.

Abstract: A method for storing a membership function, include storing a position of a vertex of a triangle that defines the membership function in a universe of discourse and storing a first distance between the position of the vertex a point of intersection between a left side of the triangle and an axis of the universe of discourse. Further, the method includes storing a second distance between the position of the vertex and point of intersection between right side of the triangle and the axis of the universe of discourse. The present invention furthermore relates to a circuit for calculating a grade of membership of an antecedent of a fuzzy rule, and is adapted to fuzzify an input variable by adopting the geometric proportions that occur between homologous sides of similar triangles defined by the position of the input value in the universe of discourse.

Abstract: A direct adaptive fuzzy controller having an adaptation mechanism that modifies the locations of output membership functions to improve performance of the fuzzy controller. The controller has a fuzzy rule base that has a different or unique output membership function for each fuzzy rule. The adaptation mechanism modifies the location of the output membership functions in response to the performance of the controller system, to continuously improve its performance. The controller is a feedback mechanism that functions not only using measured outputs of the process controlled, but has feedforward compensation that causes the controller to anticipate feedback due to measured disturbances or other parameters of the process.

Abstract: A circuit (14) and method which determine a degree of membership of an input in a membership set during a fuzzy logic operation. The degree of membership is calculated by a single "MEM" software instruction. The MEM instruction determines whether the system input has a degree of membership of zero, of a saturation level, or of some value in between. If a membership function has a boundary with an infinite slope, the slope value is set equal to zero and the degree of membership is then set equal to a saturation value for those input values corresponding to the infinite slope boundary.

Abstract: The arrangement stores, in a knowledge base memory (KBM), important (dominant) and less important rules such that they can be distinguished from one another. From a rule weighting (gi) and from an aggregated rule weighting (Gi-1) is formed. The unsharp OR linking has a unit (BSUM) for forming a limited sum, if an important rule is present, and has a unit (MAX) for forming a maximum if a less important rule is present. In this way it is ensured that each important rule, for which the rule evaluation results in a rule weighting greater than zero, has an effect on the sharp value of the relevant output variable.

Abstract: A membership function unit includes a membership function setting section for adjusting positions or shapes of membership functions depending on inputs of elements influencing human senses so as to set the membership functions and a membership value computing section for computing membership function values associated with input variables in the membership functions attained by the membership function setting section. The invention also relates to a fuzzy control system development support apparatus. The invention also relates to a method of developing a fuzzy control system.