[in Japanese]

View full abstract

[in Japanese]

View full abstract

[in Japanese]

View full abstract

[in Japanese]

View full abstract

[in Japanese]

View full abstract

[in Japanese]

View full abstract

[in Japanese]

View full abstract

With the revision of the safety law for everyday consumer goods of the fire accident is shown. Because the pressure valve operation gas of the aluminum electrolytic capacitors of the subject is white smoke, a case registered as a fire accident appears. The pressure valve operation gas is a steam mist of a safe electrolyte. I report that I have little influence on human body and report that aluminum electrolytic capacitors are not dangerous part. In addition, I introduce that there is technology to prevent ignition and consider a handling method at the time of the pressure valve operation.

View full abstract

To identify anomalous conditions in the electric power apparatus, a variety of diagnosing methods such as the nueral networks and the decision tree have been used. In this paper, we have investigated the AdaBoost for its diagnosis accuracy. Adopting the neural networks as the weak learner, the diagnosis accuracy seems to be improved. Applying the test data to the rules obtained by the training data, however, we suspect that the classifier meets the over-fitting. We have to be cautious in case of using the high-performance classifier.

View full abstract

We discuss a rule for selection of prior distributions in a hierarachical Bayes model for inference of binomial parameters. Attention is particularly paid to prior distributions defined in [0, 1] such as the beta and logistic-normal distributions. We show that when employing the beta distributions, inference of the binomial parameters and hyperparameters is sensitive to selection of hyperprior distributions and that on the other hand, when employing the logistic-normal distribution, a Bayesian update strongly emerge on the inference and consequently, the inference becomes robust against the selection of the hyperprior distributions. We conclude that in order to obtain a significant Bayesian update for the binomial parameter inference in the hierarchical Bayes model, the logistic-normal prior distribution, rather than the beta prior distribution, should be chosen.

View full abstract