CS6381_Adversarial_Distributed_Machine_Learning

This package is built for CS6381 Distributed Systems (Vanderbilt) final project, constructed for adversarial distirbuted machine learning.

Collaborators

Ran Hao (rxh349@case.edu) Tong Liang (liangtong39@gmail.com) Xiaodong Yang (xiaodong.yang@vanderbilt.edu)

• Please contact us if there is any problem

Dependencies

• ZeroMQ: http://zeromq.org/
• Java: sudo add-apt-repository ppa:webupd8team/java ; sudo apt update ; sudo apt install oracle-java9-installer ; sudo apt-get install openjdk-9-jre-headless
• Zookeeper: Download the Zookeeper package from official website: http://www.gtlib.gatech.edu/pub/apache/zookeeper/ ; Uncompress: tar xvzf zookeeper-'version'.tar.gz
• Kazoo: pip install Kazoo
• Pandas: sudo pip install pandas
• Scikit-learn: sudo pip install scikit-learn

Before running any nodes:

• First go to the zookeeper directory and cd /bin then run the zookeeper servers by : ./zkServer.sh start

Run one batch of worker nodes:

• Go to the package directory
• chmod +x worker_batch.sh
• ./worker_batch.sh scheduler_id This will run the scheduler and its corresponding worker nodes.

Run server node:

• Go to the package directory
• python server_node.py

Run individual worker node:

Each worker node belongs to a mini-batch, where a scheduler is in charger of noticing the server node of the liveness of the worker

• Go to the package directory
• python worker_node.py "worker_id" "port(default 5560)" "scheduler_id"

Run individual scheduler node:

• Go to the package directory
• python scheduler.py "scheduler_id"