Abstract

Complex network phenomena – such as information cascades in online social networks – are hard to fully observe, model, and forecast. In forecasting, a recent trend has been to forgo the use of parsimonious models in favor of models with increasingly large degrees of freedom that are trained to learn the behavior of a process from historical data. Extrapolating this trend into the future, eventually we would renounce models all together. But is it possible to forecast the evolution of a complex stochastic process directly from the data without a model? In this talk I show that model-free forecasting is possible. I present SED, an algorithm that forecasts process statistics based on relationships of statistical equivalence using two general axioms and historical data. SED is, to the best of my knowledge, the first method that can perform axiomatic model-free forecasts of complex stochastic processes. In the talk I also present promising results of simulations over simple and complex evolving processes and tests performed on a large real-world dataset.