Researches on Qualitative Recognition of Flushed Zone in Q Block on Deep Forest
Rui Su · IOP Conference Series Earth and Environmental Science · 2019
Abstract In this paper we propose a method on qualitative recognition of flushed zone. We extract the morphometric parameters of logging curves to better reflect the change relations. We use deep forest model which learns the representation layer by layer. The model is more robust and have a higher accuracy than other model. This method has an excellent application performance in Q block.