Fourier KAN network

Well logs are interpreted/processed to estimate the in-situ reservoir properties (petrophysical, geomechanical, and geochemical), which is essential for reservoir modeling, reserve estimation, and production forecasting. The modeling is often based on multi-mineral physics or empirical formulae. When sufficient amount of training data is available, machine learning solution provides an alternative approach to estimate those reservoir properties based on well log data and is usually with less turn-around time and human involvements.