Seeding with Costly Network Information

In the presence of contagion, decision makers strategize about where in a network to intervene (e.g., seeding new product). A large literature has developed methods for approximately optimizing choice k seeds cause largest cascade of, example, product adoption. However, it is often impractical measure an entire social network. “Seeding with Costly Network Information,” Eckles, Esfandiari, Mossel, and Rahimian develop analyze algorithms making bounded number queries then selecting seeds. They prove hardness results this problem provide almost tight approximation guarantees their proposed under widely used models contagion. One algorithm practical both querying online networks structuring in-person surveys. This framework further allows reasoning tradeoffs between spending budget on collecting more data versus increasing