Abstract: The vacuum cleaner includes a housing (2) having a dust compartment (11) containing a dust bag (17) and a vacuum compartment (12) containing a motor-ventilator set (20); a vacuum hose (3, 5) coupled to a suction unit (7) with a suction aperture; and a device (30) for automatically controlling the power of the motor-ventilator set (20), comprising a sensor (50) for the type of flooring and a fuzzy logic control unit (46). The automatic control device (30) of the invention further includes a sensor (60) for detecting the dynamic movement of the suction unit (7), wherein the fuzzy logic control unit (46) is adapted to apply a fuzzy inference operation to the output of the floor-type sensor (50) and to the output of the suction unit dynamic movement sensor (60), so as to control the motor-ventilator set (20) on the basis of the fuzzy inference operation.

Abstract: In a system comprising a plurality of resources for performing useful work, a resource allocation controller function, which is customized to the particular system's available resources and configuration, dynamically allocates resources and/or alters configuration to accommodate a changing workload. Preferably, the resource allocation controller is part of the computer's operating system which allocates resources of the computer system. The resource allocation controller uses a controller neural network for control, and a separate system model neural network for modelling the system and training the controller neural network. Performance data is collected by the system and used to train the system model neural network. A system administrator specifies computer system performance targets which indicate the desired performance of the system.

Abstract: A system, method, and product provide rapid explanations for the scores determined by a neural network for new observations input into the neural network. The explanations are associated with a table of percentile bins for each of the input variables used to define the observation. The table contains for each input variable a number of percentile bins. Each percentile bin contains an expected score for values of the input variable containing with the percentile bin. The expected score in each percentile bin is determined from historical observation data. Preferably each percentile bin is associated with an explanation that describes the meaning of the value of the input variable falling within the percentile bin. During observation processing, a new observation is scored. The value of each input variable in the new observation is compared with the percentile bins for the input variable in the table.

Abstract: A fuzzy neural network system includes a learning function. The learning is performed by determining degrees of coincidence of rules from combinations of membership functions for realizing fuzzy rules, constructing a network based on the number of input and output items, in such a manner as to produce output in conformity with the degrees of coincidence, to thereby properly simulate relations between input and output as to sample data represented by a subject input pattern and an output pattern corresponding thereto. The system further includes a fuzzy rule setup portion for programming fuzzy rules created by engineers into the fuzzy neural network, a fuzzy rule extracting portion for extracting each fuzzy rule from the network after a learning period, and a degree-of-importance extracting portion which extracts, for each input, a respective contribution ratio of each input on each output in the network, after a learning period.

Abstract: In a process an apparatus for operating a fuzzy inference processor in which it is verified, in sequence for each input variable, whether the conditions of a number of rules are fulfilled, and in which for one input variable in each case only those conditions of rules are verified whose conditions for the preceding input variable had been met in the method and apparatus a screening device having an address memory, an address formation unit and a device control signal former operate in combination for detecting rule segment values which continue to be relevant, thereby achieving the advantage of rapid rule evaluation.

Abstract: A method for behavioristic-format coding of quantitative resource data for control, planning and communication purposes in line businesses, utilities, etc. An individual coding unit can be used in a specific real-time object-oriented behavioristic control loop, and provide coherent, flexible data for control, planning, budgeting and forecasting, using a digital Tough Stochastic Decomposition (TSD) principle as the general behavioristic method. In accordance with the TSD method, the input raw data is converted in a standard rectangular digital matrix format, in which the 24-hour calendar day is divided into four 6-hour intervals. The coding apparatus is connected to primary instrumentation and provides total distributed raw data group coding, in non-programmable fashion, providing four digital-coded numbers a day, as four time history files.

Abstract: Circuit (SPSA) has a processing unit (VE) with an electrically erasable and electrically programmable read-only memory (EEPROM), a classification unit (KE) and an interface unit (EIF) for the read-only memory. The processing unit (VE) reads out from the read-only memory (EEPROM) instructions and/or data which have been automatically programmed in by a classification unit (KE) and an interface unit (EIF), as a function of input values (E) and/or internal values (ED, S) and/or output values (A) of the processing unit (VE). The processing unit may in this case take the form of a fuzzy controller or else a customary microprocessor. The advantage achieved hereby is, in particular, that the arrangement automatically adapts itself to the variance in structure (type diversity) and variance in time (ageing, wear) of a product and consequently the circuit has a wider range of applications and can be used optimally for a longer period of time.

Abstract: A method of automatically controlling and verifying telecommands in a satellite control system in which a satellite status analyzing/processing unit and a telecommand producing/executing unit are closely connected with each other to share a knowledge base with information regarding the telecommands. The present method comprises the steps of transmitting the telecommands to a satellite for the control thereof, receiving the resultant telemetry from the satellite, analyzing the received telemetry, inferring telemetry values corresponding to the transmitted telecommands from the information in the knowledge base, verifying a telecommand execution status of the satellite on the basis of the analyzed result and inferred telemetry values and producing a control command upon recognizing an abnormal status of the satellite in accordance with the verified result. According to the present invention, an operator needs not check data one by one to determine whether the telecommand execution is normal or not.

Abstract: A method and system for processing a plurality of fuzzy logic rules. The system includes a plurality of fuzzy logic lines, each fuzzy logic line corresponding to one of the fuzzy logic rules and including a calculating device. Each calculating device has an input terminal for receiving a series of weights and an output terminal for outputting an overall truth value according to the received series of weights and at least one logical operator of the fuzzy logic rule corresponding to the fuzzy logic line. The system further includes processing circuitry coupled to each fuzzy logic line, for receiving the overall truth value from each line, and outputting a fuzzy logic value according to the received overall truth values.

Abstract: In a method for supplementing missing data in a time series used as an input to a neural network or for improving noise-infested data supplied to a neural network, error distribution densities for the missing values are calculated on the basis of the known measured values from the time series and their known or predetermined error distribution density, and samples are taken from this error distribution density according to the Monte Carlo method. These each lead to an estimated or predicted value whose average is introduced for the value to be predicted. The method can be employed for the operation as well as for the training of the neural network, and is suitable for use in all known fields of utilization of neural networks.

Abstract: The disclosed system provides an improved collaborative filtering system by utilizing a belief network, which is sometimes known as a Bayesian network. The disclosed system learns a belief network using both prior knowledge obtained from an expert in a given field of decision making and a database containing empirical data obtained from many people. The empirical data contains attributes of users as well as their preferences in the field of decision making. After initially learning the belief network, the belief network is relearned at various intervals when additional attributes are identified as having a causal effect on the preferences and data for these additional attributes can be gathered. This relearning allows the belief network to improve its accuracy at predicting preferences of a user. Upon each iteration of relearning, a cluster model is automatically generated that best predicts the data in the database.

Abstract: The subject of the application relates to an arrangement comprising a rule decoder, a rule evaluation unit and a unit for forming selection signals, to which arrangement, from a fuzzification circuit, a minimum and a maximum number of relevant linguistic values of a respective input variable and values . . . me of membership functions of linguistic values whose numbers lie between the minimum and maximum number, and from a knowledge base memory numbers for linguistic values, prescribed in a plurality of rules, of the respective input variable, can be fed, and processing of these fed numbers and values can be executed simultaneously in parallel with processing of the respective next input variable in the fuzzification circuit. An advantageous refinement of the subject of the application also permits defuzzification running partly simultaneously in parallel with the rule decoding and evaluation.

Abstract: In a system comprising a plurality of resources for performing useful work, a resource allocation controller function, which is customized to the particular system's available resources and configuration, dynamically allocates resources and/or alters configuration to accommodate a changing workload. Preferably, the resource allocation controller is part of the computer's operating system which allocates resources of the computer system. The resource allocation controller uses a controller neural network for control, and a separate system model neural network for modelling the system and training the controller neural network. Performance data is collected by the system and used to train the system model neural network. A system administrator specifies computer system performance targets which indicate the desired performance of the system.

Abstract: A temporal learning neural network includes a plurality of temporal learning neural processing elements and an input/output control section. Each element includes a calculation device and a learning device. The calculation device includes an input memory section and a response calculation circuit. The learning device includes a learning processing circuit and a history evaluation circuit. The calculation circuit calculates a sum of a total summation value of a product of input values and connection efficacies, and an internal potential, compares the sum with a predetermined threshold value, outputs a 1 or 0 signal depending on the comparison and substitutes internal potential of a next time for the sum. The processing circuit receives an input history evaluation value when the calculation circuit has produced an output 1 signal which strengthens, weakens or leaves unchanged the connection efficacies depending on the comparison.

Abstract: A system and method forecast geomagnetic events and resulting currents from ground and space weather data, including solar wind velocity data and interplanetary magnetic field data. The system has a processor including a first prediction generator for predicting a midnight equatorial boundary (MEB) value; a second prediction generator for predicting a polar cap potential (PCP) value from the ground and space weather data; an AL and AU prediction generator for predicting AL and AU values; a pseudo Kp value generator for generating a pseudo Kp related value; an electric field pattern generator for determining electric field patterns from the pseudo Kp value, the PCP value, and the ground and space weather data; a conductivity generator for determining conductivity values from the ground and space weather data and the pseudo Kp value; and an adaptive feedback generator for adaptively generating the geomagnetic parameters from the conductivity values, the electric field values, and the predicted AL and AU values.