Automatic Discrimination of Acoustic Emission Generated in Plants from Environmental Noises using the Random Forest

DOI Web Site

Bibliographic Information

Other Title
  • ランダムフォレストを用いた植物起源弾性波とノイズ波の自動判別
  • ランダムフォレスト オ モチイタ ショクブツ キゲン ダンセイハ ト ノイズハ ノ ジドウ ハンベツ

Search this article

Abstract

<p>In recently years, acoustic emission (AE) method has been proposed to detect drought-induced stress as a biological information in plants. For the practical application of the AE method to measure drought-induced stress in plants, it is necessary to extract AE generated in plants (burst-type AE) from a large number of environmental noises. In this paper, automatic discrimination between burst-type AE and noises is attempted using decision tree and random forest classifiers. Both decision tree and random forest classifiers could discriminate between them with an accuracy of more than 0.85. In particular, the random forest classifier with many explanatory variables mitigated overlearning and discriminated the burst-type AE more accurately.</p>

Journal

Related Projects

See more

Keywords

Details 詳細情報について

Report a problem

Back to top