Application based on semi-supervised learning algorithm to land evaluation
Yueju Xue · Jisuanji gongcheng yu sheji · 2008
In order to improve the facility, interpretability and accuracy for the land evaluation model, and to reduce extended human influence in traditional land evaluation model, a land evaluation method based on semi-supervised learning algorithm is proposed.Extracting land evaluation association rules by training a small amount of labeled samples as the supervised information, combining with the use of the unsupervised learning method K-mean algorithm, the method clusters a great amount of unlabeled samples, which takes advantage of implication facilityand high classification accuracy.Experimental results of Guangdong Provinceland resource demonstrate that, 94.0622% correct area rate of land evaluation is obtained by the semi-supervised learning algorithm.