This repository contains a machine learning project that predicts the Frames Per Second (FPS) performance of various games based on a system's CPU, GPU, and game configuration. The model leverages hardware specs and setting information to forecast gaming performance, useful for gamers, system builders, and hardware reviewers.
Dataset link - FPS Benchmark - Uploaded by Ulrik Thyge Pedersen
- 🔍 Data Preprocessing: Cleaned and stripped byte-string artifacts, handled missing values, and converted categorical variables using one-hot encoding.
- 🧠 Feature Engineering: Extracted CPU/GPU brands, dropped irrelevant or redundant columns, and normalized the FPS target variable.
- 📊 Exploratory Analysis: Correlation analysis performed to find the most influential features on FPS.
- ⚙️ Modeling Techniques:
- Linear Regression
- Random Forest Regressor
- XGBoost Regressor
- 🧪 Hyperparameter Tuning: GridSearchCV used for optimal parameter selection.
The dataset (fps_benchmark.csv) contains system hardware specifications and corresponding FPS results across various games and settings.
Features include:
CpuBrand,GpuBrand- GPU architecture, memory type, interface, DirectX/OpenGL/Vulkan support
- Game titles and settings
- CPU clock speeds and process size
Install the required libraries:
pip install pandas numpy matplotlib seaborn scikit-learn xgboost