The Dynamic Relationship between Macroeconomy and Stock Market in China: Evidence from Bayesian Network

Table 1

Hill-Climbing algorithms.

---

(1) Choose a network structure over node set , usually empty.

(2) Compute the score of , denoted as .

(3) Set .

(4) Repeat the following steps as long as increases:

 (a) For every possible arc addition, deletion, or reversal not resulting in a cyclic network,

  (i) compute the score of the modified network , ;

  (ii) if , set and .

 (b) Update with the new value of .

(5) Return the directed acyclic graph .

---