FIFA-World-Cup-2022

DolphinScheduler machine learning "FIFA World Cup 2022" betting workflow

YouTube: Predict FIFA World Cup 2022 champion and get a daily betting strategy by DolphinScheduler!

In this project, we will use machine learning to predict which country will win the FIFA World Cup 2022 and get a betting strategy every day.

World Cup predictions

Here are the results of two different forecasting methods

Simulate 1000 World Cup qualification results based on team win probability

Probability of championship

Brazil       0.193
Argentina    0.160
France       0.140
Spain        0.137
England      0.131

The top four

No.1: Brazil
No.2: France
No.3: Argentina
No.4: Belgium

Choose the team with a high probability of winning

No.1: Argentina
No.2: France
No.3: Spain
No.4: Belgium

All competition results can be viewed in the following two files

https://github.com/jieguangzhou/FIFA-World-Cup-2022/blob/workflow-pro/results/predict.txt

The above results are from the branch workflow-pro. This branch will train longer models and run more simulations

Building a prediction system in three steps

Step-1 start DolphinScheduler

we can start a Dolphinscheduler standalone server using Docker

docker run --name dolphinscheduler-standalone-server -p 12345:12345 -p 25333:25333 -d jalonzjg/dolphinscheduler-fifa

And then, you can log in to the DolphinScheduler at http://localhost:12345/dolphinscheduler/ui

user: admin password: dolphinscheduler123

Step-2 submit workflow

python3 -m pip install apache-dolphinscheduler==3.1.1

export PYDS_HOME=./
python3 pyds.py

You can click the Project -> FIFA

Then, we can see 3 workflow

• training: Use FLAML to train model
• predict: Use the model to predict which country will win the World Cup
• betting-strategy: Get betting strategy every day

Step-3 run workflow

start training workflow

We can view log after the workflow had finished

Start predict workflow

We can view log after the workflow had finished

Start betting strategy workflow

$[yyyy-MM-dd]mean dolphinscheduler will use the current year, month and day as a parameter, we can also set it to 2022-12-01 or other date.

Some Notes

• Can not download data in download_data task, please check Can not access data in some region

• Can not see task status

• get_odds task error