How can Trump win? KusmartsevaAnna F. ZhangWu ZhangXinyue KusmartsevFeodor 2018 In this paper, the McCulloch-Pitts model built on an artificial neuron is first introduced briefly, followed by a modified model – the coupled network model to describe social opinion network in period of the presidential election. To illustrate the new model, its formalism and analytical results on fixed points will be stated step by step. Then, we investigate the dependence on the ratio of the initial conditions so that we could find out more on relationship between current information and preference on final results. Finally, U.S. election campaign in 2016 will be examined comprehensively including support rates, possible preference, time series analysis, and period analysis. Besides mathematical research, we also take real-life activities into consideration. For example, Trump used Twitter to help his view spreading and take advantage of the underlying uncertainty to some extent.