FIFA_Clustering

Introduction

In the given notebook, I decided to conduct a exploratory data analysis on the results of different matches in FIFA Cup. The approach I went for was performing clustering on the data to see if the team perform similarly regardless of the team they're playing against. I perform this clustering under the assumption that the statistics provided in the dataset are a measure of the performance of the team.

Analysis Steps

The steps performed in this analysis:

• Reading Data
• Feature Engineering
  • Finding Missing Values
  • Categorical Encoding
  • Feature Correlation (Barplot, Heatmap)
• Kmeans, PCA and t-SNE
  • Finding the optimal number of clusters (Elbow Method)
• Plotting the clusters
  • using annotations for different teams
  • using colours for different clusters
  • using cluster centers to determine variance between clusters