Development of Pedo-transfer Functions for Estimating Soil Aggregation and Erodibility in Kandi Region of Punjab, India

Quantification of soil aggregation and erodibility from easily measurable characteristics have been done by using pedo-transfer functions (PTFs) PTFs developed were compared statistical machine learning techniques for the kandi region Punjab. Dataset 1, having six basic properties, was used estimation mean weight diameter (MWD) (K), prediction an artificial neural network (ANN) slightly better than a generalized linear model (GLM). In dataset 2, properties in 1 high correlation with parameters GLM ANN. 3 including all 11 ANN significantly GLM. Thus, performs complex system greater number variables whereas small set fewer variables, methods perform better.