/fifabaysian

FIFA Prediction using naive bayes

Primary LanguageJupyter Notebook

Using Bayesian methods of Machine Learning to predict the outcome of the World Cup 2018

The historical results for the games played by nations over several tournaments and friendly matches was obtained from kaggle.com/martj42/international-football-results-from-1872-to-2017. For this project, we only considered the games played after 1990.

RESULTS

Country Win Prob
Panama 3.40782532144e-21
Tunisia 3.33870889796e-17
Costa Rica 1.06054606769e-15
Iceland 1.15047168436e-15
Peru 9.49614627603e-15
Iran 4.08477523177e-13
Australia 5.47302353197e-13
Japan 8.92773312939e-12
Saudi Arabia 1.1233077319e-11
Korea Republic 1.77807982253e-10
Egypt 4.84056461218e-10
Nigeria 6.971036813e-09
Morocco 2.81934931466e-08
Russia 4.39508383554e-08
Mexico 4.64058794879e-08
Sweden 5.89491079778e-08
Serbia 1.16489197847e-07
Colombia 3.24268552996e-07
Senegal 2.89937494555e-05
Denmark 0.000485088548152
Switzerland 0.000500315725254
Poland 0.000868767325681
Croatia 0.000929753461322
Uruguay 0.00958912600344
England 3.69990985263
Portugal 3.77546986039
Belgium 7.85135736102
Germany 10.0409017291
Brazil 11.5166605718
Argentina 13.0688666473
Spain 14.0985860463
France 35.9358452608

Results for the group stage below -- Fill in the scores yourself!