A system and method for yield prediction in machine learning based semiconductor manufacturing are provided. The result prediction method according to one embodiment of the present invention is a method of learning a neural network model by classifying different types of data according to types and inputting them to different neural network models, and classifying the input data into types. The result value is predicted by inputting data into different neural network models in accordance with the classification according to. This makes it possible to apply a different neural network model according to the type of data and to accurately predict a result value by securing a neural network model having a structure that matches the characteristics of the data. [Selection diagram] FIG.
展开▼