Artificial neural network modeling in bird behavior and reactions to environmental parameters in Wajiro tidal flat reclamations

Lorene L. Abella, Ryo Akasaka, Narumi Shikasho, Tatsuya Matsumoto, Koji Morita, Kenji Fukuda

研究成果: Contribution to journalArticle査読

抄録

Reclamation projects give an impression that a severe environmental damage is certain. However, the extent of damage cannot be measured by a rule of thumb. In this paper, an analysis of the environmental conditions focusing on the population of certain bird species was performed. It is only natural to surmise that bird population is decreasing due to man-made structures, much more destroying a natural habitat, though, at this stage of the study the degree as to how much the population has changed remains unknown. A model of the biological brain, known as artificial neural networks (ANN), which is the main essence of this research, might open doors to a more rigorous investigation of the environmental changes that are occurring due to tidal flat reclamations. This network is able to train itself from input parameter values and thus can predict values of desired output variables. A specific type of ANN algorithm used for calculation is the back propagation algorithm, which is also known as the generalized delta rule. Input parameters like air temperature, daylight hours, and tidal flat organisms that birds feed were chosen. Sensitivity analysis was performed to identify the birds' behavioral patterns and their reaction to the state variables.

本文言語英語
ページ(範囲)211-226
ページ数16
ジャーナルMemoirs of the Faculty of Engineering, Kyushu University
63
4
出版ステータス出版済み - 12 1 2003
外部発表はい

All Science Journal Classification (ASJC) codes

  • エネルギー(全般)
  • 大気科学
  • 地球惑星科学(全般)
  • 技術マネージメントおよび技術革新管理

フィンガープリント

「Artificial neural network modeling in bird behavior and reactions to environmental parameters in Wajiro tidal flat reclamations」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル