Abstract: A multi-user system for creating and maintaining a knowledge base is implemented on a variety of computer systems, including single-user personal computers, networked personal computers, and data communications networks, including the Internet. The system has a graphic user interface which can be easily learned by non-programmers. The knowledge base is displayed as a hierarchical multiplicity of nodes, in which each node is related to one or more superior nodes. Title of text, rules, references, and attributes associated with each node are displayed, and may be modified. Simulation is used to verify the knowledge base once generated. Data may be extracted from the bases in the form of hard-copy books, databases for downloading into Web browsers, and databases for downloading into hand-held computers including a multiplicity of grids, and a multiplicity of branches in the form of a logic tree.

Abstract: Matching (e.g., via correlation or similarity process) entities having attributes, some of which have associated values. The values of the attributes may be adjusted based on number of entities that have values for a particular attribute so that the values decrease as the number increases. The attributes of the entities may be harmonized and provided with default values so that entities being matched have common attributes defined by the union of the attributes of the entities being matched. The attributes of the entities may be expanded and provided with default values so that the entities being matched have attributes that neither had originally. Match values may be normalized to provide a weight value which may be used to predict an attribute value of a new entity based on known attribute values of known entities. The weight values may be tuned such that relatively high weights are amplified and relatively low weights are suppressed.

Abstract: A system and method are disclosed for partitioning a real-value windowed attribute into ranges, wherein the values within each range generally correspond to a particular class of results associated with runs of a process. The system and method determines a low range having attribute values generally corresponding to a first class, a middle range having attribute values generally corresponding to a second class, and an upper range having attribute values generally corresponding to the first class. The system and method may be used in a system that produces an induction tree useful in developing an indication of a cause of a particular result of a process from values associated with at least one real-valued, windowed attribute that arises during the runs of the process.

Abstract: The fuzzy filtering of a noise signal comprising a plurality of signal samples [s(t,k)] is carried out using as variables the variation of the signal in the considered window and the distance of the samples from a sample to be reconstructed, to distinguish the typical variations of the original signal from those due to the noise and to identify the signal fronts. The method comprises the steps of defining a current signal sample [s(t)] from among the plurality of signal samples, calculating a plurality of difference samples [D(t,k)] as the difference in absolute value between the current signal sample and each signal sample and defining distance values (k) between the current signal sample and each signal sample.

Abstract: A system for interfacing a human user to a data processor which receives inputs from the user and includes associated storage resource information. The data processor generates outputs to the user and associated output devices. The interface system includes structure for receiving a statement generated by the human user in natural language on a word-by-word basis. The system analyzes the statement to identify a subject and object and searches the stored resource information for data related to the identified subject. Output devices are provided for generating to the user the data from the stored resource information related to the identified subject.

Abstract: An input data group is converted into an output data group in accordance with coefficients. For each coefficient, a current coefficient value, a coefficient value greater than the current coefficient value and a coefficient value smaller than the current coefficient value may be computed. One of the current, greater and smaller coefficient values is selected on the basis of errors between the output data computed with the current, greater and smaller coefficient values and an ideal output data group corresponding to the input data group.

Abstract: A system and a method for tracking long term performance of a vibrating body such as a gas turbine, includes a vibration sensor who time domain outputs are transformed to the frequency domain, using a fast Fourier transform processing. Frequency domain outputs are provided as inputs to a fuzzy adaptive resonance theory neural network. Outputs from the network can be coupled to an expert system for analysis, to display devices for presentation to an operator or are available for other control and information purposes.

Abstract: The analog processor of this invention is programmable and capable of storing the processing coefficients in analog form. It comprises a storage section having at least one output, plural outputs in most cases, and being adapted to respectively generate programming signals on such outputs; the storage section is input a plurality of supply voltage signals and is operative to produce, in connection with information stored therein, one of the supply voltage signals on each of the outputs, it being understood that one voltage signal may be produced on several such outputs. Advantageously, the processor can also be programmed in a simple manner from circuits of the digital type if switches controlled by storage elements are used in the storage section.

Abstract: A “virtual string” is generated for synthesizing sound produced by plucked-string instruments using recurrent neural networks. The disclosed recurrent neural network, called a Scattering Recurrent Network (SRN), is based on the physics of waves traveling in the string. Vibration measured from a plucked string is used as the training data for the SRN. The trained SRN is a virtual model capable of generating tones similar to the tones generated by the physical string. As with a real string, the “virtual string” corresponding to the SRN responds differently to different types of string “plucking” motions.

Abstract: Fully automated methods and systems for processing complex data sets to identify abnormalities are described. In one embodiment, the system includes wavelet processing, recursive processing to determine prominent features, and then utilizing feed forward neural networks (FFNNs) to classify feature vectors generated in the wavelet and recursive processing. With respect to wavelet processing, multiresolution (five-level) and multidirection (two-dimensional) wavelet analysis with quadratic spline wavelets is performed to transform each image. The wavelets are a first-order derivative of a smoothing function and enhance the edges of image objects. Because two-dimensional wavelet transforms quantize an image in terms of space and spatial frequency and can be ordered linearly, the data is processed recursively to determine prominent features. A neural network approach derived from sequential recursive auto-associative memory is then used to parse the wavelet coefficients and hierarchy data.

Abstract: A fuzzy logic information processing system and method are disclosed that may be used not only with known or definite data input but also with uncertain data input. The uncertain data input may be represented by a set of values wherein the possibility of any particular or specific value within the set being the true or accurate value is uncertain. The preferred embodiment of the system provides for an extensor to extend or map a representation of the uncertain data into at least one additional dimension related to dimensions of a set of rules used for making fuzzy logic inferences. The set of rules may be provided effectively in a mapped or graphed form. The set of rules and uncertain data are combined, for instance by locating intersection regions, to produce an output set that may be also be described as a map or plot. In a presently preferred embodiment the uncertain inputs and rules are represented mathematically or symbolically and then operated on to produce an output set.

Abstract: A method and apparatus is disclosed for machine learning of a pattern sequence which is derived from a plurality of inputs. The pattern sequence is predicted from learning rate parameters that are exponentially related to an incrementally calculated gain parameter for each input. The gain parameter are increased or decreased in real time in correlation with the accuracy of the learning process. The disclosed method and apparatus are advantageously utilized in signal processing, adaptive control systems, and pattern recognition.

Abstract: A system for detecting interaction between different services on a telecommunications network includes a computer expert system. A data store in the expert system is programmed with data which represent attributes of service features. A rule store is programmed with rules which relate feature attributes to interaction behaviors. An inference engine is connected to the data store and to the rule store and processes the data and the rules to detect any interaction between the services. The data in the data store may be arranged as sets of objects, each object in a set corresponding to a different state transition of the corresponding feature. The different objects may be given sequence numbers corresponding to the time sequence of execution of the feature. At least some of the rules may relate to these sequence numbers.

Abstract: In this invention there is disclosed a system and method for generating a fuel-optimal reference velocity profile for a rail-based transportation handling controller. The fuel-optimal reference velocity profile is used to reduce the amount of fuel consumption. The fuel-optimal reference velocity profile is generated to take into account the speed that a train must travel at in order to satisfy a given schedule, speed constraints that will minimize fuel consumption, and recommended speeds that will prevent train breaks, derailments, cargo damage and violation of safety rules. In this invention, the fuel-optimal reference velocity profile is generated by using a genetic algorithm.

Abstract: A method for learning to classify data according to two distinct classes (c11, c12) separated by a separating surface (S), by means of a neurone of the binary type comprising a parameter describing the separating surface and whose inputs are weighted by a weight (wi), and including the following steps: a) defining a cost function C: C σ = ∑ μ = 1 ( γ μ > o ) P ⁢ [ A - B ⁢   ⁢ tanh ⁢   ⁢

Abstract: Described are method and apparatus for discriminating whether a time series data is a deterministic chaos or a stochastic chaos. In the method and apparatus, a trajectory parallel measuring method is used to discriminate the observed time series data.

Abstract: A system which observes the world through a video camera and/or other sensors, automatically learns a probabilistic model of normative behavior through the use of a Hidden Markov Model, and uses that model to infer the kind of activity currently under view and detect unusual behavior. The use of hidden Markov models is made possible by entropic training of the model with an &thgr;&thgr; entropic estimator that folds structure learning into the parameter estimation process to remove parameters from the Hidden Markov Model which have little information content, thus to permit real time robust unusual behavior detection. In one embodiment, the system consists of three components: image analysis; model learning; and signal analysis. In image analysis, each frame of video is reduced to a vector of numbers which describe motion of objects in front of the camera, with a sequence of such vectors, one for each frame of video, establishing the “signal.

Abstract: Motor vehicle sensor signals are evaluated by a fuzzy system, which generates control signals for a system device of the motor vehicle—for example an automatic transmission, active suspension, speed stabilization, power-steering assistance, or traction control. The fuzzy system is connected to a neural network, which evaluates the sensor signals and reference data from a recording of driving data of the motor vehicle. The neural network optimizes the rule base of the fuzzy system. During a driving operation, the fuzzy system generates on-line signals categorizing the respective driving situation, and thus makes possible intelligent, time-adaptive, driving-situation-dependent control. The fuzzy system and the neural network each contain a classification system which can be reciprocally converted by a correspondence-maintaining bidirectional transformation.

Abstract: The invention relates to a modular architecture of a cellular network for improved large-scale integration, of the type which comprises a plurality of fuzzy cellular elements (C.sub.m,n) interconnected to form a matrix of elements having at least m rows and n columns, the row and column numbers describing the location of each element. Each fuzzy processor is adapted for connection to other processors of the same type such that a parallel architecture of the modular type can be implemented. The management of the architecture is facilitated by each submatrix being controlled by an individually dedicated fuzzy processor device.

Abstract: A system is disclosed that provides a goal based learning system utilizing a rule based expert training system to provide a cognitive educational experience. The system provides the user with a simulated environment that presents a business opportunity to understand and solve optimally. Mistakes are noted and remedial educational material presented dynamically to build the necessary skills that a user requires for success in the business endeavor. The system utilizes an artificial intelligence engine driving individualized and dynamic feedback with synchronized video and graphics used to simulate real-world environment and interactions. Multiple "correct" answers are integrated into the learning system to allow individualized learning experiences in which navigation through the system is at a pace controlled by the learner.