Matsushita Electric Industrial Co., Ltd.

University of California

Kansai Paint Company, Ltd.

Rion Company Ltd.

Technology Research Association of Medical and Welfare Apparatus

Hideyuki Takagi

Tatsumi Watanabe

Susumu Maruno

Kazuaki Obara

Isao Hayashi

Masahiro Watanabe

Shigeo Sakaue

Yasuharu Shimeki

Shinichi Sakamoto

Toshiyuki Kohda

Tsuyoshi Mekata

David M Auslander

Eiji Mizutani

Hayato Togawa

A parameter writing element converts a solution vector found by an optimum solution method that determines optimum n-dimensional solution vector, based on a plurality of optimum n-dimensional solution vector candidates to an adjustment parameter values of a programmable hearing aid and writes the adjustment parameter values into a hearing aid parameter memory element of the programmable hearing aid using a sound source memory element for storing a sound source, and a sound source presenting element for presenting the sound source to the programmable hearing aid.

Optimum solution method, hearing aid fitting apparatus utilizing the optimum solution method, and system optimization adjusting method and apparatus

At the start of a process, an information presenting section presents information expressed by a solution vector, while an auxiliary information presenting section presents information for facilitating comparative evaluation of multiple pieces of information being presented simultaneously. Based on these pieces of information, the user evaluates each solution vector. A model estimation executing section generates an evaluation model for an adjustment process from a user evaluation history, and thereafter, a model evaluation calculating section performs evaluation using the evaluation model. A fitness calculating section calculates fitness from the evaluation made by the user or by the model, and in accordance with the thus calculated fitness, a recombination operation section performs an arithmetical recombination operation. The above process is iteratively performed to carry out optimum adjustment of solution vectors.

Optimization adjusting method and optimization adjusting apparatus

Proportion predicting system and method of making mixture

There is disclosed a pattern outline data extraction device in which plural pattern feature vectors representing patterns contained in a given pattern distribution are input, a geometric figure containing at least one feature vector on the outer periphery thereof is repeatedly generated to delete possible feature vectors contained within the geometric figure generated and an outline of the given pattern distribution is identified with feature vectors not having been deleted. An outer surface detection device and a pattern recognition device are also disclosed as applications of the pattern outline data extraction device.

Outline pattern data extraction device for extracting outline pattern of

The adjustment of a noise suppressing apparatus inputs the voice signal with the noises being piled up in it to decide the weight coefficient of the neutral network 130 with a back propagating method so that the errors of the output signal and of the noiseless signal may become minimum, the noise suppressing apparatus where noises in a voice signal are suppressed even in a situation where the positional relation between the noise source and the voice source may often change, and noises offensive to the voice after the noise suppression may not remain, the suppressing effect is not deteriorated even if the time pattern of the input signal is varied.

Noise suppressing apparatus and its adjusting apparatus

An inference rule determining process according to the present invention sequentially determines, using a learning function of a neural network model, a membership function representing a degree which the conditions of the IF part of each inference rule is satisfied when input data is received to thereby obtain an optimal inference result without using experience rules. The inventive inference device uses an inference rule of the type 'IF . . . THEN . . .' and includes a membership value determiner (1) which includes all of IF part and has a neural network; individual inference quantity determiners (21)-(2r) which correspond to the respective THEN parts of the inference rules and determine the corresponding inference quantities for the inference rules; and a final inference quantity determiner which determines these inference quantities synthetically to obtain the final results of the inference. If the individual inference quantity determiners (2) each has a neural network structure, the non-linearity of the neural network models is used to obtain the result of the inference with high inference accuracy even if an object to be inferred is non-linear.

Inference rule determining method and inference device

An iterative learning machine uses, as a direction of changing iterative weight, a conjugate gradient direction in place of the conventional steepest descent direction, thereby saving time. Learning rates are set dynamically. Error calculation for plural learning rates, with respect to a certain weight changing direction, are accomplished by storing a product-sum of the input signals and weights in a hidden layer and a product-sum of the input signals and the weight changing direction in the hidden layer. When the learning falls into a non-effective state where further iteration does not effectively reduce an error, the weights are adjusted in order to restart the learning.

Neutral network with plural weight calculation methods and variation of

Growing community of inventors

Kyoto, Japan

Hideyuki Takagi