2022 FIFA WORLD CUP PREDICTION

Using pytorch to build a neural network to predict the 2022 World Cup

Give the probability of winning and losing for each game separately

The training data comes from kaggle (https://www.kaggle.com/datasets/martj42/international-football-results-from-1872-to-2017)

Instructions

1. Download the data from the above kaggle link and decompress it, and place the results.csv in the ./data directory

2. python preprocess.py

3. python main.py

4. The output is in ./data/pred.csv

Here are the predicted results

	home_team	away_team	home_team_loss	home_team_win	home_team_draw
0	Qatar	Ecuador	0.43832436203956604	0.3241387605667114	0.23753690719604492
1	Senegal	Netherlands	0.8146077990531921	0.08299645781517029	0.1023956760764122
2	England	Iran	0.10977872461080551	0.7393050789833069	0.150916188955307
3	United States	Wales	0.4844847321510315	0.3573998510837555	0.1581154316663742
4	Argentina	Saudi Arabia	0.06608234345912933	0.7922183275222778	0.14169932901859283
5	Mexico	Poland	0.4614976644515991	0.32846954464912415	0.21003274619579315
6	Denmark	Tunisia	0.2634960114955902	0.5287051200866699	0.20779883861541748
7	France	Australia	0.26930108666419983	0.6184151768684387	0.11228371411561966
8	Germany	Japan	0.1238890290260315	0.7331917881965637	0.14291918277740479
9	Spain	Costa Rica	0.14859357476234436	0.6584973335266113	0.1929091364145279
10	Morocco	Croatia	0.6307979822158813	0.1604074239730835	0.20879463851451874
11	Belgium	Canada	0.09586668759584427	0.8012700080871582	0.10286331921815872
12	Switzerland	Cameroon	0.26291173696517944	0.46560153365135193	0.271486759185791
13	Brazil	Serbia	0.17897723615169525	0.6597122550010681	0.16131047904491425
14	Uruguay	South Korea	0.2594752907752991	0.4751227796077728	0.2654018700122833
15	Portugal	Ghana	0.23299959301948547	0.5963274836540222	0.17067290842533112
16	Qatar	Senegal	0.3303745687007904	0.42403876781463623	0.24558669328689575
17	Netherlands	Ecuador	0.09825852513313293	0.7737862467765808	0.12795525789260864
18	Wales	Iran	0.4309306740760803	0.4346366822719574	0.13443270325660706
19	England	United States	0.14756451547145844	0.6854586005210876	0.16697683930397034
20	Poland	Saudi Arabia	0.16117233037948608	0.6589977741241455	0.179829940199852
21	Argentina	Mexico	0.2075396180152893	0.5802081227302551	0.21225225925445557
22	Tunisia	Australia	0.3612756133079529	0.3422435522079468	0.29648083448410034
23	France	Denmark	0.5277354121208191	0.379800021648407	0.09246461093425751
24	Japan	Costa Rica	0.39247211813926697	0.39483439922332764	0.2126934677362442
25	Spain	Germany	0.50408536195755	0.3060823678970337	0.18983221054077148
26	Belgium	Morocco	0.2701772153377533	0.5029365420341492	0.22688625752925873
27	Croatia	Canada	0.04068133234977722	0.8759750127792358	0.08334363996982574
28	Cameroon	Serbia	0.6986231803894043	0.14093028008937836	0.16044658422470093
29	Brazil	Switzerland	0.14428852498531342	0.6458815932273865	0.2098298966884613
30	South Korea	Ghana	0.3239705264568329	0.39121779799461365	0.28481167554855347
31	Portugal	Uruguay	0.4403846859931946	0.354835569858551	0.20477978885173798
32	Ecuador	Senegal	0.4165593087673187	0.3586595952510834	0.22478114068508148
33	Qatar	Netherlands	0.788962185382843	0.08795084059238434	0.12308698147535324
34	Wales	England	0.7688230872154236	0.11704124510288239	0.11413561552762985
35	Iran	United States	0.4404911994934082	0.3711323142051697	0.18837648630142212
36	Poland	Argentina	0.621189534664154	0.180080845952034	0.19872967898845673
37	Saudi Arabia	Mexico	0.6524074077606201	0.19572623074054718	0.15186640620231628
38	Australia	Denmark	0.6347565054893494	0.22331339120864868	0.14193014800548553
39	Tunisia	France	0.5312294363975525	0.26178571581840515	0.20698480308055878
40	Japan	Spain	0.7042847871780396	0.1324021965265274	0.16331300139427185
41	Costa Rica	Germany	0.8122816681861877	0.08854364603757858	0.09917477518320084
42	Croatia	Belgium	0.28506723046302795	0.5130628347396851	0.2018699198961258
43	Canada	Morocco	0.6158958077430725	0.16862329840660095	0.21548084914684296
44	Serbia	Switzerland	0.2926972806453705	0.48347508907318115	0.22382767498493195
45	Cameroon	Brazil	0.8536257743835449	0.04531894996762276	0.10105519741773605
46	Ghana	Uruguay	0.7040308713912964	0.15104389190673828	0.1449252963066101
47	South Korea	Portugal	0.4871547520160675	0.27046167850494385	0.24238361418247223

2022 世界杯预测

使用pytorch搭建神经网络预测2022年世界杯

分别给出每场比赛的输赢概率

训练数据来自kaggle (https://www.kaggle.com/datasets/martj42/international-football-results-from-1872-to-2017)

使用方法

1. 从上述链接下载数据集并解压,把results.csv放到./data中

2. python preprocess.py

3. python main.py