Influence Learning and Maximization

The problem of maximizing or minimizing the spreading in a social network has become more timely than ever with advent fake news and coronavirus epidemic. solution to this pertains influence maximization algorithms that identify right nodes lockdown for epidemic containment, hire viral marketing campaigns, block online political propaganda etc. Though these have been developed many years, majority literature focuses on scalability issues relaxing method’s assumptions. In recent emergence new complementary data advanced machine learning methods decision guided part towards learning-based approaches. These can range from how information spreads over network, solve combinatorial optimization itself. tutorial, we aim dissentangle clearly define different tasks around applications networks. More specifically, start traditional algorithms, describe need estimation delineate state-of-the-art diffusion learning. Subsequently, delve into while optimizing which is based algorithms. Finally, latest approaches graph neural networks deep reinforcement