A Static-Dynamic Hypergraph Neural Network Framework Based on Residual Learning for Stock Recommendation

Figure 3

The illustration of hypergraph construction and hypergraph convolution. In the process of hypergraph construction, we follow the following steps. Initially, each node search for its nearest neighbor, which is marked as a star node in the figure. For the sake of clarity in our illustration, we depict two-star nodes in the figure, representing the closest points among the surrounding nodes. These star nodes are enclosed within blue dotted lines, effectively forming hyperedges. Next, we treat the star node as the central node and create a hyperedge by applying the k-nearest neighbor algorithm. The connection is denoted by a red dotted line (in the figure, it is indicated as 2-nn).