HANDBOOK OF RANDOM FORESTS: THEORY AND APPLICATIONS FOR REMOTE SENSING

This compendium provides a hands-on description of random forests. The text starts with a consistent introduction of general methods to create, train, and fuse ensembles of decision trees. Instead of limiting the explanation to the general-purpose layout of traditional random forests, this book outlines specifications during tree creation and training, that are especially well suited to analyze structured data such as images. The theoretical foundations are explained as deeply as practical and implementation issues. The many possible variations of the underlying Random Forest model are discussed as well as their implications on the outcome in order to provide insights into the influence of these parameters and their possible side-effects. Last but not least, this unique title provides specific examples of the usage of Random Forests for analysis tasks of remote sensing imagery.

Readership: Researchers, professionals, academics and graduate students in pattern recognition/image analysis, machine perception/computer vision, and artificial intelligence.

這本彙編提供了對隨機森林的實用描述。本書首先介紹了創建、訓練和融合決策樹集合的一般方法。與僅限於傳統隨機森林的通用佈局的解釋不同，本書概述了在樹的創建和訓練過程中的規格，這些規格特別適合分析結構化數據，如圖像。理論基礎的解釋既深入又實用，並涉及實施問題。討論了底層隨機森林模型的許多可能變化以及它們對結果的影響，以提供有關這些參數的影響及其可能的副作用的見解。最後，這本獨特的書籍提供了使用隨機森林進行遙感圖像分析任務的具體示例。

讀者群：模式識別/圖像分析、機器感知/計算機視覺和人工智能領域的研究人員、專業人士、學者和研究生。