숫자야구 GA

Using Genetic Algorithm to solve Bulls & Cows ⚾

About

Bulls & Cows are known as 'Number Baseball' in Korea.

The rules are the same, but we also call the in-game terms differently:

• bulls are strikes
• cows are balls

Just think about baseball and you'll get the main idea. 😁

This project uses GA to solve Bulls & Cows:

1. Generate the secret number randomly & create the first generation of genes which are scored(20pts for each strikes, 10pts for each balls).
2. Select 4 elite genes that have the most highest scores and make the next generation.
3. Crossbreed random genes until we have 6 descendant genes and add them in next generation genes(Mutations can happen with a probability of 10% while crossbreeding).
4. Switch to next generation and score the genes in the current generation.
5. Repeat step 2~4 until the exact solution(the gene with a perfect score) is found.

Example

[*] GOAL : 73461
1 : 54063(40), 62158(20), 59746(30), 84097(20), 64180(30), 02639(20), 27195(20), 63479(60), 72196(40), 64520(20)
2 : 63479(60), 54063(40), 72196(40), 59746(30), 72163(60), 63499(50), 53496(50), 72496(50), 72493(50), 59726(20)
3 : 63479(60), 72163(60), 63499(50), 53496(50), 62193(30), 63179(50), 53490(40), 73196(60), 52496(30), 53196(40)
4 : 63479(60), 72163(60), 73196(60), 63499(50), 72103(40), 72196(40), 73496(70), 72106(40), 63109(40), 73106(60)
5 : 73496(70), 63479(60), 72163(60), 73196(60), 63199(40), 73166(80), 73066(70), 73169(70), 73163(80), 73063(70)
6 : 73166(80), 73163(80), 73496(70), 73066(70), 73196(60), 73896(50), 73466(90), 73169(70), 72896(30), 53466(70)
7 : 73466(90), 73166(80), 73163(80), 73496(70), 73463(90), 78163(60), 78496(50), 78462(60), 73066(70), 78162(50)
8 : 73466(90), 73463(90), 73166(80), 73163(80), 73263(70), 73266(70), 79466(70), 78163(60), 71266(60), 78263(50)
9 : 73466(90), 73463(90), 73166(80), 73163(80), 73460(80), 73160(70), 83160(50), 83460(60), 79460(60), 78160(50)
10 : 73466(90), 73463(90), 73166(80), 73163(80), 73453(70), 63453(60), 63157(50), 63451(70), 63457(60), 73457(70)
11 : 73466(90), 73463(90), 73166(80), 73163(80), 93463(70), 73469(80), 73183(60), 73489(60), 73462(80), 73486(70)
12 : 73466(90), 73463(90), 73166(80), 73163(80), 73486(70), 73169(70), 73469(80), 75186(40), 75486(50), 23486(50)
13 : 73466(90), 73463(90), 73166(80), 73163(80), 23163(60), 23466(70), 73126(60), 23166(60), 73863(70), 73266(70)
14 : 73466(90), 73463(90), 73166(80), 73163(80), 03166(60), 73462(80), 03462(60), 79166(60), 79462(60), 13462(70)
15 : 73466(90), 73463(90), 73166(80), 73163(80), 73496(70), 71466(80), 71462(70), 73462(80), 73162(70), 71563(60)
16 : 73466(90), 73463(90), 73166(80), 73163(80), 70463(70), 13463(80), 70163(60), 73406(70), 75406(50), 75463(70)
17 : 73466(90), 73463(90), 73166(80), 73163(80), 73461(100), 73161(90), 73168(70), 73108(50), 53161(70), 73468(80)
[*] Reached 73461 in 17 generations

Average of generations needed to reach goal in each 10000 executions

1. When Selection Pressure is 0.5 with random goals

50.3837
50.1111
50.0761

2. When Selection Pressure is 0.6 with random goals

45.1001
43.9827
45.0043