Affordable Analytics on Expensive Data
Prasang Upadhyaya, Martina Unutzer, Magdalena Bałazińska, Dan Mircea Suciu, Hakan Hacígümüş · 2014
In this paper, we outline steps towards supporting "data analysis on a budget" when operating in a setting where data must be bought, possibly periodically. We model the problem, and explore the design choices for analytic applications as well as potentially fruitful algorithmic techniques to reduce the cost of acquiring data. Simulations suggest that an order of magnitude improvements are possible.