Oracle Corporation
Machine learning for optimal student guidance

Abstract:

Techniques for training and evaluating machine-learning models for providing student guidance are described herein. In some embodiments, a network service generates a set of clusters that group a plurality of students by similarity. The network service trains a machine-learning model based on variances in outcomes and actions leading to the outcomes for students that belong to a same cluster in the set of clusters. The network service evaluates the machine-learning model for a student that has been mapped to the same cluster to identify at least one action that the student has not performed that is predictive of an optimal outcome for other students that belong to the same cluster. Responsive to evaluating the machine-learning model, the network service presents, through an interface, a recommendation that the student perform the at least one action.