FootBall League Prediction Using Machine Learning

Project Overview

The objective of this project was to analyze the dataset, preprocess the data, train multiple regression models, compare their performance, and identify the best-performing model. This project was developed as part of our Machine Learning coursework/project work.

---

👩‍💻 About Me

Hi, I’m Fatima Shahid, a Computer & Information Systems Engineering Undergraduate and an enthusiastic problem solver.

• 🧠 Solving LeetCode problems for 2+ years
• 🐍 Primary language: Python
• 🏗️ Strong interest in DSA, Competitive Programming, Systematic Thinking, Operating Systems and Machine Learning
• 📈 Believer in progress through daily effort

Objectives

• Perform data preprocessing and feature engineering
• Train and evaluate multiple Machine Learning models
• Compare model performance using evaluation metrics
• Fine-tune selected models for improved accuracy
• Select the best-performing model for CEP prediction

---

Models Implemented

Decision Tree Regressor

• Depth = 3
• Depth = 5
• Depth = 10

Linear Models

• Linear Regression
• Ridge Regression
• Lasso Regression

Ensemble Models

• Random Forest Regressor
• Fine-Tuned Random Forest Regressor

---

Technologies Used

• Python
• Scikit-learn
• Pandas
• NumPy
• Matplotlib
• Seaborn
• Google Colab

---

Dataset Processing

The following preprocessing steps were performed:

• Handling missing values
• Feature selection
• Train-validation-test split
• Time-based and random splitting experiments

---

Evaluation Metrics

The models were evaluated using:

• Mean Squared Error (MSE)
• Root Mean Squared Error (RMSE)
• R² Score

---

Final Model Performance

Model	MSE	RMSE	R² Score
Fine-Tuned Random Forest	92.97	9.64	0.59

The Fine-Tuned Random Forest model achieved the best overall performance among all tested models.

---

Project Structure

├── DataFiles/
├── notebook
├── models/
├── Reports/
├── README.md

---

Installation

Clone the repository:

git clone https://github.com/FatimaaShahid/Football-Match-Outcome-Prediction-using-Machine-Learning

Run the notebook or Python files to train and evaluate the models. Don't forget to save the trained NN models in drive first.

---

Future Improvements

• Increase dataset size for better generalization
• Perform advanced hyperparameter tuning
• Experiment with deep learning models
• Apply additional feature engineering techniques
• Improve model explainability and visualization

---

Acknowledgements

We would like to acknowledge the valuable contributions and teamwork of all project members:

• Asra Siddiqui
• Muhammad Abdullah

Their support, collaboration, and efforts played an important role in the successful completion of this project.

---

Authors

• FATIMA SHAHID
• Asra Siddiqui
• Muhammad Abdullah

📬 Contact & Connect

• 💼 LinkedIn: Fatima Shahid
• 🧠 LeetCode: FatimaaShahid
• ✍️ Medium: @fatimashahid781
• 📧 Email: fatimashahid781@gmail.com
• 📸 Instagram: @fatimas_abstract

---