Apply and use data in your work! Learn how to influence decision-making through a strong command of accurate and timely data. Predictive analytics involves the use of current and historical data, statistical algorithms and machine learning techniques to identify the likelihood of future outcomes.

This three-week online course is designed to enhance your ability to apply some basic techniques to develop predictive models from data.

Introduction

Our world changes with unprecedented speed. Anticipating emerging trends and predicting the subsequent effects has become increasingly important for designing strategies that successfully address the most pressing global issues and ensure that “no one is left behind”. This course is intended to introduce some of the basic techniques to develop predictive models from data. This is an intermediate level course, that requires participants to have a reasonable understanding of the basic concepts of data analysis (see course Fundamentals of Data Analytics in the UN context).

Objectives

Upon successful completion of the programme, participants will be able to:

  • Explore data to better understand relationships among variables
  • Apply some basic techniques to develop predictive models from data
  • Choose and implement appropriate performance measures for predictive models
  • Understand how to ensemble models to improve predictions
Course methodology

This course is delivered entirely online, in partnership with IE University faculty and key experts in data analytics. It combines live webinar sessions, led by subject-matter experts, with self-paced activities and interactive group discussions.

The weekly instructor-led webinars are conducted in the Zoom online platform. The webinars will take place on June 9, June 16 and June 23, from 15:00 to 16:30 CET. Participants need a computer (or mobile device), a reliable internet connection and either headset with microphone to connect to the audio through the computer, or a telephone. We recommend accessing audio through the computer. No special software is required; but participants must be able to access Zoom. We will send instructions for Zoom access to registered participants and recommend that you download the application and test your access in advance.

The self-paced components and discussion forum for each week of the course are designed and structured on UNKampus, UNSSC’s Learning Platform.

Course contents

Week 1: Why prediction and modelling from data?

  • Predictive modelling in the UN
  • What is a predictive model?
  • Time dimension in predictive analytics

Week 2: Predictive Analytics

  • Modelling from data - applications
  • Predictive modelling - regressions
  • Machine Learning for predictive modelling

Week 3: Applying predictive analytics in intergovernmental scenarios

  • The wisdom of the crowds- ensembles and multi-modelling
  • What can and cannot be predicted, black swans and future tellers
  • Ethical issues in Machine Learning

The practices will be performed using BigML, which is a tool that has a free license for users.

Target audience

All UN personnel (professional and general service staff) at headquarters and field locations. If participants are interested in getting a full experience from the fundamentals of data analytics, to a more intermediate level of data analytics, we highly recommend to enrol simultaneously in this course as well as in the course Fundamentals of Data Analytics in the UN context.