Distributed-AI-kernel

Basic implementation of Gossip Learning (machine learning over fully distributed data), for the case of classification with linear models.

The project is a small demo implementation of Robert Ormandi and Istvan Heged's scientific paper.

Getting Started

git clone https://github.com/Tribler/distributed-ai-kernel.git

We recommend to have Maven installed on your local machine:

sudo apt update

sudo apt install maven

Finally check the installation

mvn -version

Running

mvn exec:java 

OUR_AVERAGE SLOPE: 0.7034892602092983  OUR AVERAGE INTERCEPT: 0.6037076408018176
[BEST_FIT LINE]: SLOPE:0.7142857142857143 intercept: 0.6071428571428572
BEST FIT LINE error:1.6071428571428577
OUR LINE error:1.6139083467505504
RELATIVE ERROR: 0.995808008787293

Running the tests

mvn test

Authors

• JayDew - Initial work - GitHub
• Matei Enache - Initial work - GitHub
• Johan Pouwelse - Coordonator - GitHub

License

This project is licensed under the MIT License

Acknowledgments

• Hat tip to anyone whose code was used
• The project is still in early days of development