A system and method are disclosed for training a machine learning model using information pertaining to transmissions of one or more media items to user devices associated with a user account. In an implementation, training data for the machine learning model includes first contextual information associated with a first user device and second contextual information associated with a second user device. The number of the transmissions to the user devices for the user account exceeds a threshold number of transmissions allowed for the user account. Training data further includes generating a first target output that identifies an indication of a preference of a user preference to keep or cancel each of the transmission. The method includes providing the training data to train the machine learning model. The trained machine learning model may be used to determine which of the new transmissions is to be canceled.