For full details, please read the Potluck and FoggyCache paper. Currently, our ongoing project is to explore broader applicability of approximate reuse for database and machine learning systems. Namely, combining semantic analysis techniques and approximation theories to enable semantic-level computation reuse with bounded error. Stay tuned.
Approximate computation reuse
Publications
Peizhen Guo, Bo Hu, Rui Li, Wenjun Hu.
FoggyCache: Cross-Device Approximate Computation Reuse.
In MobiCom,
2018.
Peizhen Guo, Wenjun Hu.
Potluck: Cross-Application Approximate Deduplication for Computation-Intensive Mobile Applications.
In ASPLOS,
2018.