[Hackerrank] Cut the tree

The problem is available here. The idea is to use postorder traversal of the tree to enumerate each possible remove of edge, and return the sum of the sub-tree to its parent node. This problem represent the tree as a acyclic undirected graph, which is a quite wired representation.

public class Solution {

    public static class GraphNode {
        int id, val;
        List<GraphNode> neighbors;
        public GraphNode(int id, int val) {
            this.id = id;
            this.val = val;
            neighbors = new ArrayList<GraphNode>();
        }
    }

    static int result = Integer.MAX_VALUE;

    public static int dfs(GraphNode root, int sum, HashSet<Integer> visited) {
        if (root == null) {
            return 0;
        } else {
            int curSum = 0;
            visited.add(root.id);
            for (GraphNode neighbor : root.neighbors) {
                if (visited.contains(neighbor.id) == false) {
                    int res = dfs(neighbor, sum, visited);
                    // System.out.println(sum + " " + res);
                    result = Math.min(Math.abs(sum - 2 * res), result);
                    curSum += res;
                }
            }
            return curSum + root.val;
        }
    }

    public static void main(String[] args) {

        Scanner scan = new Scanner(System.in);
        int n = scan.nextInt();
        List<GraphNode> nodes = new ArrayList<GraphNode>();
        int sum = 0;
        for (int i = 0; i < n; ++i) {
            int val = scan.nextInt();
            nodes.add(new GraphNode(i+1, val));
            sum += val;
        }
        int a = scan.nextInt(), b = scan.nextInt();
        GraphNode root = nodes.get(a-1);
        nodes.get(a-1).neighbors.add(nodes.get(b-1));
        nodes.get(b-1).neighbors.add(nodes.get(a-1));
        for (int i = 0; i < n-2; ++i) {
            a = scan.nextInt();
            b = scan.nextInt();
            nodes.get(a-1).neighbors.add(nodes.get(b-1));
            nodes.get(b-1).neighbors.add(nodes.get(a-1));
        }
        HashSet<Integer> set = new HashSet<Integer>();
        dfs(root, sum, set);
        System.out.println(result);
    }
}