Team builder model predictor

In this project I built an engine that could help Liga NOS's coaches build their team before a game. Data Collection:

• Gather data of 3 and a half seasons;
• Web scrapping zerozero.pt (Gathering which players played in each game);
• Acquire datasets on football-data.co.uk (Gathering the stats of each game).

Data Wrangling:

• Transform datasets into a suitable dataframe in order to train and test the machine learning model;
• The main dataset is the difference between the home team's stats and the away team's stats.

Caracterization of Liga NOS's teams and players: https://public.tableau.com/profile/tom.s.marques#!/vizhome/LigaNOS_16154766812000/Dashboard1

Machine Learning model (Decision tree):

• Predicts the result of a game Uses these stats to predict the result:
• Half time goals;
• Corners;
• Fouls committed;
• Shots;
• Shots on target;
• Yellow cards;
• Red cards.

Product: result_predictor_liga_nos.py

Script to gather the data: Scrap Football stats 2.ipynb

Script to build the model: model_final_project.ipynb