Fig. 3
From: Evaluating soil nutrients of Dacrydium pectinatum in China using machine learning techniques
A schematic structure of a neural network. Field measurements taken from soil nutrients (total nitrogen, organic matter content, alkali-hydrolysable nitrogen, available phosphorus, and rapidly available potassium) were used as input values provided through the input layer. Five evaluation indicators were used as dependent variables. Soil nutrient quality evaluation results are represented by an output node. Four layers of a hidden layer were constructed