A mobile cloud distributed machine learning system

• المؤلف الرئيسي: Tran, Vu Xuan Nhat
• مؤلفون آخرون: Ta Nguyen Binh Duong
• التنسيق: Final Year Project
• اللغة: English
• منشور في: 2017
• الموضوعات: DRNTU::Engineering::Computer science and engineering::Computer systems organization::Computer-communication networks
• الوصول للمادة أونلاين: http://hdl.handle.net/10356/70424
• المؤسسة: Nanyang Technological University

Distributed Machine Learning (DML) has recently become popular and Parameter Server is an easy to use and efficient framework to solve DML problems. Setting up and configuring Parameter Server are challenges to data analytics and researchers with focusing in training and testing their models. We implement a cloud web applications to integrate with an Auto Deployment Platform for Parameter Server along with a Benchmark as references. The cloud web system includes a back-end system to deal with integration and scale problems, and a cross-platform web responsive front-end interacts with end-users. To demonstrate the implementation of our web applications, we go through from client interface implementations to database, back-end API and finish with system integration. We show and explain the scenario in which available and suitable techniques are used to achieve our objectives.