RobertoFalconi/GameRatingsPredictor

Machine learning algorithm solves multi-class classification problem of video games content rating (without playing it). Quantitative Methods for Computer Science exam project.

Game Ratings Predictor

This is a project originally made by Roberto Falconi and Federico Guidi for the course "Quantitative Methods for Computer Science" and its teacher Luigi Freda, based at Sapienza - University of Rome.

SlideShow: https://www.slideshare.net/RobertoFalconi4/game-ratings-predictor-machine-learning-software-to-predict-video-games-content-rating

Useful links

• Portfolio: https://robertodaguarcino.com
• LinkedIn: https://www.linkedin.com/in/roberto-falconi
• GitHub: https://github.com/RobertoFalconi
• SlideShare: https://www.slideshare.net/RobertoFalconi4

The code is open source and written in Python 3.x but it's also Python 2.x backward compatible.

The program uses the Kaggle dataset available at https://www.kaggle.com/rush4ratio/video-game-sales-with-ratings and the Python libraries "Pandas" and "Sklearn".

This project goal is to classifie each video game in the dataset by ESRB rating, to do this we used Logistic Regression, Random Forest and k-NN.

How to run GameRatingsPredictor

Ubuntu, Debian and macOS users: First of all, you need pip and Python installed.

If you're on Ubuntu or Debian, make sure you have Python Developer installed with sudo apt-get install python-dev.

If you're on macOS, you can install Homebrew with /usr/bin/ruby -e "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install)" or update it with brew update (you can check it here: https://brew.sh) and then both Python and pip thanks to brew install python.

1. Upgrade pip to the latest version with pip install --upgrade pip or pip2 install --upgrade pip.
2. Download the repository and open a terminal inside the root.
3. Run sudo pip install -r requirements.txt or sudo pip2 install -r requirements.txt, it will install all the required packages.
4. Enter python algoritmo-runner.py to run the project with without any installation.

You're done!

In addition to this, you can install the program so with sudo python setup.py install --record files.txt. Now you can run the whole project with the command line algoritmo anywhere in the terminal. You can uninstall it with cat files.txt | xargs rm -rf.

Windows users:

1. Go to https://www.python.org/downloads/windows/, download and run the .exe

2. Download the get-pip.py file from https://bootstrap.pypa.io/get-pip.py

3. Install it with (you might need an administrator command prompt to do this): python get-pip.py

4. Go to the directory where Python has been installed, now in the Scripts directory make a new file called local.bat with the only word cmd in it

5. Double click local.bat and in the just opened terminal write:

    pip install numpy
    pip install scipy
    pip install pandas
    pip install -U scikit-learn
   

6. Run algoritmo.py with Python IDLE or the IDE you prefer!