Day 1 - Networks

Practical info

Place: DHØ 1.23 Time: 8:15 (we start 15 min later) - 13:20

Schedule for the day

TimeActivityData
Session 18:15-9:00Intro , Q&A, Introducing datasetDanish power elites
Session 29:00-11:30Network explorationDanish power elites
Session 312:00-13:15WS: AI as a service

Dataset & Context

Introduction

Context: The Danish Power Elites

Data

Tasks

  • Who are the most central persons?
  • Communities?
  • What characterizes them?
    • Link up with additional data?

Workshop: AI as a Service

By guest: Andreas Markussen

Introduction

The purpose is to demonstrate a minimum viable product of an ML model deployment. Keeping with the DSBA spirit, I would keep the section hands-on. I would cover a minimal ML pipeline using data that is already preprocessed. The model will be saved to a static file and served through an API. When done, we will have created a web service from scratch, that can be used by any user in their browser. The project will be conducted in Python, with a little HTML added.

Structure:

  • Very brief introduction to how the web works - 15 mins
  • Building model and exporting to static file - 15 mins
  • Building Flask web app - 10 mins
  • Adding APIs - 10 mins
  • Adding some HTML and concluding the complete web service - 10 mins
  • Deployment to Heroku free account, short showcase of extensions and other applications (no code here) and rounding off - 15 mins

Preperation

To get most of this workshop, you should have:

  • Some existing knowledge of Python
  • A local installation of Python
  • Good understanding of an ML flow.
  • No understanding of the web in general

Materials