Industrial Technology Research Institute

Ming-Chun Hsyu

Chien-Chih Huang

Wen Tsuei

Chao-Nan Chen

Wen-Pin Hsu

Chun-Te Yu

Chao-Hong Chen

Chao-Hung Chen

A data processing method used in neural network computing is provided. During a training phase of a neural network model, a feedforward procedure based on a calibration data is performed to obtain distribution information of a feedforward result for at least one layer of the neural network model. During the training phase of the neural network model, a bit upper bound of a partial sum is generated based on the distribution information of the feedforward result. During an inference phase of the neural network model, a bit-number reducing process is performed on an original operation result of an input data and a weight for the neural network model according to the bit upper bound of the partial sum to obtain an adjusted operation result.

Data processing method and device for bit-number reduction in a neural network by establishing bit upper bound and bit lower bound using distribution information

A method and an electronic device for selecting deep neural network hyperparameters are provided. In an embodiment of the method, a plurality of testing hyperparameter configurations are sampled from a plurality of hyperparameter ranges of a plurality of hyperparameters. A target neural network model is trained by using a training dataset and the plurality of testing hyperparameter configurations, and a plurality of accuracies corresponding to the plurality of testing hyperparameter configurations are obtained after training for preset epochs. A hyperparameter recommendation operation is performed to predict a plurality of final accuracies of the plurality of testing hyperparameter configurations. A recommended hyperparameter configuration corresponding to the final accuracy having a highest predicted value is selected as a hyperparameter setting for continuing training the target neural network model.

Method and electronic device for selecting deep neural network hyperparameters

A method for determining the accuracy of the estimated position for a target device in a wireless system includes the computation of a confidence index. In the online location determining phase, after knowing the observations of the radio signal for a target device, the target device's probability distribution of location and its motion model are combined to calculate the position uncertainty, thereby giving the confidence index of this location estimate. The invention determines the location probability distribution, and calculates the uncertainty of the location probability distribution and the possible maximum uncertainty under the current situation. Based on these uncertainties, this invention determines the confidence index of the radio signal. The confidence may be regarded as a quantity that the location uncertainty can be excluded in the location prediction.

Method and apparatus for determining accuracy of the estimated location for a target in a wireless system

Disclosed is a method and apparatus for determining the new sample points of the location determination system in a wireless environment. In the training phase, after knowing the distribution of the observed radio signal of a location for a target device, the uncertainty of the probability distribution of the received radio signal is analyzed by a location probability distribution model. The radio signal distribution differences are calculated, and the penalties between the location in question and its nearby locations are also calculated, thereby determining whether a location is the candidate of new sample points or the recalibration points. This invention has nothing to do with decision rules and will not undergo the problems of choosing inappropriate decision rules. It also provides a penalty model for generating penalty of error prediction from one location to another location, which can as well easily mitigate the problem on the sample points with the double-role attribute of the border line.

Growing community of inventors

Taichung, Taiwan

Ming-Chun Hsyu