Game Popularity Prediction

Project Description

To predict the popularity of games using modern machine learning techniques, we begin with comprehensive preprocessing to ensure data quality and relevance. This involves handling null values, removing duplicates, and encoding complex data types such as dates and lists. Additionally, new columns are created based on existing ones to enhance the dataset's informative value.

Both regression and classification models are applied to the dataset to capture different aspects of the prediction problem. The algorithms employed include:

1. Linear Regression: For predicting continuous popularity scores.
2. Logistic Regression: For binary classification tasks.
3. K-Nearest Neighbors (KNN): For instance-based learning to capture similarity patterns.
4. Support Vector Machine (SVM): For finding the optimal hyperplane in classification tasks.
5. Decision Tree: For intuitive, tree-based predictions.
6. Random Forest Classifier: For ensemble learning, leveraging multiple decision trees to improve accuracy and robustness.

Project Team

• Seif Omran
• Seif Samer
• Seif Ezz
• Seif Wael
• Shawkat Sayed