Generic web-service client for cloud-based machine learning platforms

Mehmet Ali Rende

Master's thesis

Abstract:

The thesis consists of a full-stack web service application: the client-side application written with JavaScript / React Framework, the server-side application written with Python / Django REST Framework. The server listens to input and interactions on the user interface and sending responses back after analyses of the input. These two applications serve as a web service that compares different behaviors of cloud-based machine learning platforms and tests them from different aspects with input from the user interface in order to evaluate predictive performances in real-time and ease potential users’ decision process. On the other hand, the thesis discusses possible use cases and future adaptions of the application. The thesis includes the full implementation of the demo application, business logic process flows as well as the documentation explaining exactly how the web service implemented and works actively. Additionally, the paper includes findings from two cloud-based machine learning platforms’ comparison.

Abstract:

The thesis consists of a full-stack web service application: the client-side application written with JavaScript / React Framework, the server-side application written with Python / Django REST Framework. The server listens to input and interactions on the user interface and sending responses back after analyses of the input. These two applications serve as a web service that compares different behaviors of cloud-based machine learning platforms and tests them from different aspects with input from the user interface in order to evaluate predictive performances in real-time and ease potential users’ decision process. On the other hand, the thesis discusses possible use cases and future adaptions of the application. The thesis includes the full implementation of the demo application, business logic process flows as well as the documentation explaining exactly how the web service implemented and works actively. Additionally, the paper includes findings from two cloud-based machine learning platforms’ comparison.

Keywords

Machine learning, Python, REST, Cloud systems, API, Web service

Keywords

Machine learning, Python, REST, Web service, API, Cloud Systems

Language used: English

Date on which the thesis was submitted / produced: 8. 10. 2019

Thesis defence

• Date of defence: 13. 1. 2021
• Supervisor: Tomáš Kliegr
• Reader: David Chudán

Vysoká škola ekonomická v Praze

Master programme / field:
Aplikovaná informatika / Information Systems Management