Disclosed is a method, apparatus, and device for scheduling virtual objects in a virtual environment, relating to artificial intelligence, and falling within the technical field of computers. The method comprises: acquiring frame data generated by an application program in a virtual environment during operation; Performing feature extraction on the frame data to obtain state features of the target virtual object; Inferring state features to obtain N types of subsequent state features; Calling the value network prediction model to process N types of subsequent state features, thereby obtaining expected returns of execution of the N types of scheduling policies; And controlling the target virtual object to execute a scheduling policy having the highest expected revenue.
展开▼
