Predictive Analysis of FIFA World Cup Using Machine Learning

The task of predicting the outcome of the FIFA World Cup is a complex challenge due to its tactical nature and player-specific variables. In this paper, we propose a machine learning framework to predict match results using a dataset of World Cup matches played between 2002 and 2022. We applied feature engineering to extract predictive metrics such as recent team form, Head-to-head historical performance, FIFA rankings, and position-specific indicators (e.g., goalkeeper rankings). We trained and compared two complementary classifiers, Logistic Regression (LR) and Random Forest (RF), and performed hyperparameter optimisation to address class imbalance. The results showed that the Random Forest model outperformed Logistic Regression with a higher AUC of 0.7830.