Efficient Feature Embeddings for Student Classification with Variational Auto-encoders

Jun 25, 2017·
S. Klingler
Dr. Rafael Wampfler
Dr. Rafael Wampfler
,
T. Käser
,
B. Solenthaler
,
M. Gross
Abstract
We present an approach for efficient feature embeddings for student classification using variational autoencoders (VAEs). In educational data mining, student performance data is often heterogeneous, high-dimensional, and partially observed. VAEs learn compact latent representations that capture the underlying structure of student learning behavior, enabling downstream classification of student types and learning trajectories with improved generalization compared to hand-crafted features or standard autoencoders. We evaluate our approach on educational datasets from tablet-based learning environments and demonstrate improved classification accuracy with lower computational cost.
Type
Publication
In Proceedings of the International Conference on Educational Data Mining (EDM), Wuhan, China
publications
Educational Data Mining Variational Autoencoders Student Classification Feature Learning
Dr. Rafael Wampfler
Authors
Dr. Rafael Wampfler
Senior Researcher & Lecturer

I am a Senior Researcher & Lecturer at the Computer Graphics Laboratory of ETH Zurich, and a Research Consultant at Disney Research. I am leading the Digital Character AI projects at CGL. My research interests include conversational digital characters, affective computing, human-computer interaction, and applied machine learning.

My vision is to create intelligent digital humans that can naturally communicate, understand, and support people across domains such as education and mental health. My research focuses on multimodal artificial intelligence for interactive digital humans, developing models that combine large language models, affective computing, and data-driven animation to create embodied conversational agents endowed with autonomous agency, consistent values, and beliefs.

My work bridges machine learning, human–computer interaction, and computer graphics to enable AI systems such as Digital Einstein and interactive patient avatars for psychotherapy training and health education.