A Higher-Order Motif-Based Spatiotemporal Graph Imputation Approach for Transportation Networks

Algorithm 1

Motif-based graph aggregation.

Input: the node set, target node , the eigenvalue of the first associated node.

Output: imputed value .

Part I: Motif-based search.

1: Initialize: current node set , motif gain

2:repeat

3: while do

4: update

5: set according to Equation (2)

6: obtain according to Equation (1)

7: obtain according to Equation (3)

8: i ← i +1

9:end while

10:until no neighbor node exists

11: update target node set

Part II: Aggregation

12: Initialize:

13: for do

14: update according to Equations (4)–(12)

15: end for

16: return