Master's Thesis: Learning Progressively Smaller Models for Mobile Networks via Self-Distillation
Core
Design and implement a self-distillation framework using tabular foundation models to adapt to data drift in mobile networks while reducing model size and latency.
Role type
Master's thesis researcher (AI/ML)
Builds
Self-distillation pipeline for mobile network tasks
Domain
Telecommunications / AI Research
Deliverable
research
Required skills
Python, PyTorch, Machine Learning, Neural Networks, Model Compression, Knowledge Distillation
Preferred skills
Tabular Foundation Models, Continual Learning, Git
Responsibilities
Review research on tabular foundation models and knowledge distillation; Design a self-distillation pipeline; Implement and evaluate the framework on real-world datasets; Analyze trade-offs between accuracy, adaptation speed, and computational cost; Present findings in a thesis.
