Player-Recommendation-System

Link To The App

Motivation Behind The Project

Imagine your favourite team lost its most important player through a transfer, who are the players that can replace him/her ? who posseses similar XG, XA and similar style of play ? Thats one of the main questions that fans and teams always think about. A model that can show you the most similar players to your chosen player as rapidly as possible in real time

Aim of The Project

The goal here is to make a recommendation engine that lets you find players of similar style and stats to a specific selected football player

Project Timeline

• Scraped data from fbref
• Separated the data based on various playing possitions
• Performed Exploratory Data Analysis and performed Feature Engineering on these datasets
• Used Principal Component Analysis for dimensionality-reduction
• Cosine similarity was used to create the recommendation system
• Deployment was done through Streamlit and Heroku

Stills From The App