Deep learning at scale

Project Details

Project Lead
ibrahim hallac 
Project Manager
ibrahim hallac 
Institution
Firat University, Computer Science Department  
Discipline
Computer Science (401) 
Subdiscipline
14.09 Computer Engineering 

Abstract

In this project we aimed to investigate and test the current deep learning tools and libraries on cloud computing platforms. Experiments will be made by analyzing image and text data using deep learning techniques.

Intellectual Merit

There is a big potential in using deep learning technologies on big data. We believe that improving computational performance and analyzing techniques will bring benefits to the field.

Broader Impacts

We will learn and do experiments in using deep learning techniques over cloud computing platforms. Deep learning solutions generally need a lot of computational power but it's results are usually successful and better than the traditional machine learning algorithms. Increasing performance and success of the deep learning techniques and tools is an interesting topic and we aim to do the experiments by using computer clusters on the cloud and run deep learning jobs in parallel.

Scale of Use

We need about 20 VMs to run experiments. Experiments will be done in about once a week.