HIGHER-ORDER NETWORK

HON is more accurate in representing the true movement patterns in data in comparison with the conventional first-order network or the fixed second-order network, because the higher-order nodes and edges in HON can provide more detailed guidance for simulated movements... While ships from Shanghai to Singapore are in fact more likely to go to Los Angeles according to the raw data, the conventional first-order network captures only pairwise connections in the data and neglects higher-order dependencies, therefore subsequent network analyses that rely on the network structure can potentially lead to incorrect results or conclusions... The higher-order network preserves higher-order dependencies in the network structure, therefore movements simulated on the network more accurately reflects the true movement patterns in the raw data.