AISiBoCo Knowledge Base

Context of this documentation

This documentation is written in the context of the TETRA project AISiBoCo - funded by the flemish agency for innovation and entrepreneurship (VLAIO) - on the topic of deploying Machine Learning algorithms on single board computers. It aims to provide an overview of the main contents of the project in the form of a knowledge base. In the project the main differences, advantages and disadvantages of edge computing and cloud computing were discussed in the context of a healthcare scenario and a smart city scenario. In this knowledge base, the main metrics that play an important role in the decision process of a solution in the cloud or on the edge are summarized and explained. Furthermore the main ecosystems for Machine Learning using cloud services and single board computers are also detailed in this documentation.

Contents:

• Cloud, Edge and Fog Computing
  • Cloud Computing
  • Edge Computing
  • Fog Computing
• Decision metrics for Cloud VS Edge
  • Timing aspects
  • Data aspects
  • Cost Aspects
  • Privacy and security aspects
• Typical Cloud Service Providers
  • Introduction
  • Different Cloud Ecosystems for Computing
  • Amazon Web Services (AWS)
  • Microsoft Azure
  • Google Cloud Services
  • IBM Watson
  • Choice of Cloud service provider
• Typical Single Board Computer Ecosystems for Edge Computing
  • Existing Ecosystems for SBC hardware
  • NVIDIA
  • Intel
  • ARM
  • Google TPU
• Software Repositories and Project Demos
  • Summary
  • Seminar 1: Outlier detection on KDD Cup dataset
  • Seminar 2: Image and video processing for object detection and classification
  • Seminar 3: Image segmentation and model optimization
• References