The Centre hosted a two-day workshop from 11-12 April 2019 that brought together 30 people from 15 organizations to discuss the promise and challenges of using predictive analytics in humanitarian response.

The main objectives were to exchange information about predictive analytics initiatives and to identify gaps, challenges and opportunities related to the application of predictive models in crises. [Read the full report here.]

During the workshop, six organizations presented on their predictive work. The World Bank shared an overview and preliminary results from the Artemis pilots, which is based on a model that is used to predict famine risk. UNHCR shared preliminary results from Project Jetson, an effort to predict population movement from Somalia into south-eastern Ethiopia. IDMC, OCHA, the Red Cross Red Crescent Climate Centre and Save the Children also presented on their predictive models.

In a video message to workshop participants, the Under-Secretary-General for Humanitarian Affairs Mark Lowcock said, “One of the biggest opportunities we have is to try to use data, and especially the tools of predictive analytics to get ahead, to be more anticipatory, to predict what is about to happen and to trigger the response earlier.”

Below are key takeaways from the workshop:

  • There is a strong willingness to collaborate across organizations and with subject matter experts to help address current capacity gaps, and ensure models are well developed and used to support decision making and action.
  • There is agreement that models are tools, not solutions and for a model to perform well, decision makers need to be involved from the beginning (and throughout) in order to frame the question(s) for the model to answer.
  • Most of the models that were shared by organizations are in a pilot phase and still need further validation and feedback before they can be used to create a trusted signal for the sector to respond to. Until then, pilot model outputs should work alongside existing analysis and forecasts to determine what action is needed.
  • There is a desire to develop documentation and case studies of pilot models for collective learning.
  • There is particular interest in establishing a peer review mechanism to strengthen ethical deliberation and offer improved transparency and accountability in this area.
  • The sharing of model-ready data between trusted partners can save many months spent on data cleaning.

As an immediate next step, the Centre has created a modeling group for the Horn with interested organizations given the likely drought and related increase in food insecurity in the region. The Centre is also hiring a lead for its predictive analytics programme and will be considering what services it should provide such as a peer review network, ethical guidance, the development of case studies, and new models to predict humanitarian needs.

Participants share their thoughts on predictive analytics in humanitarian response in this brief workshop video.

The workshop would not have been possible without the additional support of the municipality of The Hague. Very special thanks to Deputy Mayor Saskia Bruines, The Hague’s alderman for Education, Knowledge Economy and International Affairs, for making remarks to the group on 11 April.

To get updates on the Centre’s work on predictive analytics, click here to join the mailing list on this topic. We are happy to get feedback. Send any questions to comments to centrehumdata@un.org or reach us on Twitter @humdata.

 

Predictive Analytics Workshops