Presentation 2000/5/11
One the Prediction of the Best Listening Level with Neural Network : Effect of Data Selection for Learning
Hironori Ueda, Katsuyuki Hagiwara, Yuichi Noro, Kazuhiro Kuno,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this report, we investigated influence which the data selection for training of the neural network exerted on the prediction of the best listening level. Seven neural networks were trained by using various combinations in three data sets(the best listening levels under various noises)where the group of subjects and the experiment day were different. The results suggensted that the prediction error for unlearned data is reduced if the noises in the training data set have various spectrum patterns. Moreover, twelve subjects listened the broadcasting speech, which the volume was changed according to the predicted level of the network trained by all data sets, with actual noise in a soccer stadium. As a result, the subjects of about 60 percent answered that the speech was listened comfortably.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) public address system / neural network / fluctuating noise / broadcasting level / prediction
Paper # EA2000-5
Date of Issue

Conference Information
Committee EA
Conference Date 2000/5/11(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Engineering Acoustics (EA)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) One the Prediction of the Best Listening Level with Neural Network : Effect of Data Selection for Learning
Sub Title (in English)
Keyword(1) public address system
Keyword(2) neural network
Keyword(3) fluctuating noise
Keyword(4) broadcasting level
Keyword(5) prediction
1st Author's Name Hironori Ueda
1st Author's Affiliation Faculty of Engineering, Mie University()
2nd Author's Name Katsuyuki Hagiwara
2nd Author's Affiliation Faculty of Engineering, Mie University
3rd Author's Name Yuichi Noro
3rd Author's Affiliation Faculty of Engineering, Mie University
4th Author's Name Kazuhiro Kuno
4th Author's Affiliation Faculty of Engineering, Mie University
Date 2000/5/11
Paper # EA2000-5
Volume (vol) vol.100
Number (no) 53
Page pp.pp.-
#Pages 7
Date of Issue