TY - JOUR
TI - A method for logging data reconstruction based on transfer learning
AU - Zhao Yuxuan
AU - Guo Yong
AU - Chen Hao
AU - Yang Lei
AU - Xi Chuanming
AU - Xu Yuqiang
JN - Thermal Science
PY - 2024
VL - 28
IS - 4
SP - 3451
EP - 3456
PT - Article
AB - This paper proposes a logging data reconstruction method based on migration  learning, which can reduce the dependence on labeled data and also help to  im­prove the generalization ability of data-driven models. Reconstruction  experiments are carried out using data from a block in the Junggar Basin of  Xinjiang, and compared with the conventional data-driven long short-term  memory network and recurrent neural network methods. The results show that  the reconstruction results based on migration learning improve the accuracy  by 21%, which is significantly better than the remaining two methods, and  proves the feasibility of the method.