Skip to content

Algorithms -- Greedy algorithms

Minimum spanning trees

Here are some property:

  1. Removing a cycle edge cannot disconnect a graph
  2. A tree on n nodes has n-1 edges
  3. Any connected, undirected graph with |E| = |V| -1 is a tree
  4. An undirected graph is a tree iff. there is a unique path between any pair of edges

And cut property: Suppose edges X are part of a MST of G=(V,E). Pick any subset of nodes S for which X does not cross between S and V-S, and let e be the lightest edge across this partition. Then X \cup {e} is part of some MST

Kruskal's MST algorithm

procedure kruskal(G, w){
// Input: A connected graph with weights w
// Output: A MST defined by edges X
for all u \in V:
    makeset(u);

X={};
sort the Edges E by weight;
for all edges (u,v) \in E. in increasing order of weight:
    if find(u) != find(v):
        add edge (u,v) to X;
        union(u,v)

}

So we need a kind of data structure to support such operations. It is disjoint sets

procedure makeset(x){
    pa[x] = x;
    rank[x] = 0;
}

function find(x){
    while x != pa[x]:
        x = pa[x];

    return x;

procedure union(x,y){
    rx = find(x)
    ry = find(y)
    if rx == ry: return
    if rank[rx] > rank[ry]:
        pa[ry] = rx;
    else:
        pa[rx] = ry;
        if rank[rx] = rank[ry]:
            rank[ry] = rank[ry] +1;

Path Compression

function find(x):
    if x != pa[x]:
        pa[x] = find(pa[x])
    return pa[x]

Prim's algorithm

x ={ }
repeat until |X| = |V| - 1:
    pick a set S \subset V for which X has no edges between S and V-S
    let e \in E be the minimum-weight edge between S and V-S
    X = X \cup {e}

Huffman encoding

The goal is find a method to encode the message as compressive as possible. And in fact here is

cost of tree = \sum_{1}^{n} f[i] * (depth of ith symbol in tree)

And we can also express the sum as frequencies of all leaves and internal nodes, except the root

So we can construct the whole tree from bottom by combining two nodes which are least frequent

procedure Huffman(f){
// Input: An array f[1..n] of frequencies
// Output: An encoding tree with n leaves

let H be a priority queue of integers, ordered by f
for i = 1 .. n: insert(H,i)
for k = n+1 .. 2n-1:
    i = deletemin(H), j = deletemin(H)
    create a node numbered k with children i,j
    f[k] = f[i] + f[j]
    insert(H,k)

Horn formulas

This topic is actually about logic. But how we test a satisfying assignment is related to Greedy Method.

Set Cover

This is about approximation factor i.e. even greedy method doesn't behave well, but we'd like to know how far is it from