The present invention relates to a device and method for predicting productivity of a shale gas well in a transition flow region by using a machine learning technique and, more specifically, to a device and method for predicting productivity of a shale gas well in a transition flow region by using a machine learning technique, wherein a data-based artificial neural network model is generated to calculate future production of a shale gas well in a transition flow region, and future production behaviors and estimated ultimate recovery are calculated using a decline curve method on the basis of the calculated future production and existing production data of the shale gas well in the transition flow region, whereby accurate results can be derived with less error than conventional productivity prediction methods for shale gas wells.
展开▼