Applying data and machine learning to scale education
summary
Coursera applies large-scale data and ML to make course recommendations for 24 million registered users, tackling cold-start with IP-derived country, referrer and other implicit features plus an expanded onboarding questionnaire about career goals. Courses (≈2,000) are first grouped into 100 clusters via a course co-enrollment embedding, then a two-layer classifier—decision trees at each level—selects specific courses. This pipeline raises relevance and yields a higher click-through rate compared with the previous system.
AI-written summary · about
related talks
25mMachine learning algorithms as tools for student success prediction
26mFrom Data to Pathways: Personalizing Learning Experiences
35mA Course on Machine Learning for Software Developers
10mEmpowering Personalized Learning with Big Data
27mRav Ahuja: Co-founder & Chief Program Manager at Cognitive Class – an IBM initiative at Cypher 2017
19mAdvancing Personalized Learning Analytics with AI and EdTech
44mTransforming education with AI
33mTeaching ML and AI to Coders