Deterministic Query Approximation
Three recent papers on query approximation,
- DAQ: A New Paradigm for Approximate Query Processing (Potti and Patel, 2015)
- Efficient Approximate Query Answering over Sensor Data with Deterministic Error Guarantees (Brito et al., 2017)
- BlinkDB: Queries with Bounded Errors and Bounded Response Times on Very Large Data (Agarwal et al., 2013)
have proposed new methods for approximating queries with deterministic error bounds, but there techniques have not yet been applied extensively to broader query contexts.
The goal of this project is to expand the techniques described to the TPC-H benchmark queries. In doing so, we hope to demonstrate the viability of these methods for query approximation in a wide range of possible queries.