IEEE Global Humanitarian Technology Conference (GHTC)
Technology for the Benefit of Humanity // Villanova University, USA / October 23-26, 2024

Applied Machine Learning for Social Good Workshop

Thursday,October 29, 9am-Noon PT


  • Rakshit Agrawal, Camio Inc.
    Organizer of Applied ML for Social Good Tutorial at GHTC 2019
  • Charles Delahunt, Global Health Labs, Bellevue, WA, University of Washington, Applied Math Dept


  • Noni Gachuhi, Shannon Kuyper (Global Health Labs, Global Development Technologies Portfolio)
  • Craig Nakagawa (Global Good, Partner Development)
  • Rachel Millin (Global Health Labs, ML team)
  • Charles Delahunt
  • Sourabh Kulhare (Global Health Labs, ML team)


Learning Objective

This session will focus on how to effectively apply Machine Learning (ML) methods for social good, with examples drawn from health care applications for Low and Middle Income Countries (LMICs). We will discuss a user-centric framework that includes: (i) working with field partners to identify projects and define performance specs; (ii) building partnerships to ensure successful deployment; (iii) ML considerations such as tailoring ML metrics to the use-case.

The workshop is 3 hours:

  1. Introduction to the concepts of developing ML solutions for social good applications.
  2. Several invited talks (~20 minutes) on key topics, along with Q/A.
  3. Further resources and discussion as wished

Invited Talks

1. Identifying and defining projects: Noni Gachuhi, Shannon Kuyper (Global Health Labs, Global Development Technologies Portfolio)

a. Working with field partners to identify target problems
b. Researching appropriate product performance specs
c. User-centric design and iterative feedback from the field
d. Example: Cervical cancer screening in remote settings

2. Commercialization: Craig Nakagawa (Global Good, Partner Development)

a. Receptiveness to ML in LMIC – high willingness to adopt
b. Importance of engagement by local groups and partners for successful deployments. Example: health care clusters in LMIC.
c. Partnering with the private sector for large-scale deployment

3. Case study: Lung Ultrasound: Rachel Millin (Global Health Labs, ML team)

a. Overview of the GHL lung ultrasound project, including
b. use-cases
c. commercial partners and their different product goals
d. data, data collection, and annotation
e. ML models
f. some results

4. Tailoring metrics to guide ML development: Charles Delahunt (Global Health Labs; U. of Washington, Applied Math)

a. Limitations of standard ML metrics (e.g. AUC)
b. Importance of product performance specs (as opposed to generic metrics)
c. Tailoring/defining the metrics of model outputs to the particular task
d. Examples: Malaria detection; grain moisture prediction.

5. Introduction to Deep Neural Nets: Sourabh Kulhare (Global Health Labs, ML team)

a. Introduction
b. Some system details (data, annotation, architectures, etc)
c. Examples
d. Tips-observations-pitfalls of set-up and training



We hope that the participants will acquire: (i) a systematic framework for effectively applying ML to societal challenges; (ii) key principles for effective deployments grounded in concrete examples; and (iii) a sense of ML’s role in the framework.


Instructor Bios:

Global Health Labs (Bellevue, WA) is a research lab that develops innovative solutions to address unmet needs in primary health care centers and the last mile​ in Low and Middle Income Countries (LMIC). It is funded by Bill Gates.

Rakshit Agrawal is an Applied Scientist at Camio working on Machine Learning and Computer Vision. He completed his PhD from the University of California, Santa Cruz with a dissertation on “Generalized Learning Models for Structured Data”. His research interests span AI, Machine Learning, Crowdsourcing, HCI, ICTD and AI for Social Good. He also developed and taught the UCSC course on Applied Machine Learning for Social Good.

Noni Gahuchi has a public health background and spent 16 years living and working across Africa and Asia, designing and delivering health programs for reproductive health, malaria, and HIV prevention. In her current role at Global Health Labs, she oversees a portfolio of health technology products intended for use at primary health care in low resource settings.

Shannon Kuyper leads a product development team at Global Health Labs that is focused on sound strategy support to assist in translation of ideas to products, with an emphasis on areas such as early technology/user experience and iterative field/prototype evaluations prior to handoff to manufacturing partners.  Previously she worked at Philips in New Product Introduction for ultrasound, and her educational background is in Chemistry.

Craig Nakagawa is an impact-driven technology executive with over 20 years of experience developing and launching global health technologies and services in low and middle-income countries. Currently, he is advising the World Health Organization on how technology can support COVID vaccine introductions.  He is also a Senior Operating Partner with Bamboo Finance, an Swiss-based impact fund investing in low and middle-income markets.

Rachel Millin has a background in medical imaging, neuroscience, and machine learning.  She is currently an Associate Scientist at Global Health Labs, where she develops machine learning products that address medical challenges in LMICs.

Charles Delahunt has applied Machine Learning to health care needs in LMICs for the past 7 years, and is part of the ML team at Global Health Labs. He also does basic research on Machine Learning methods, in the Applied Math department at the University of Washington.

Sourabh Kulhare is a Research Scientist at Global Health Laboratories applying Machine Learning and computer vision algorithms to different modalities of heath care data. He completed his Master’s at Rochester Institute of Technology, NY with the thesis “Deep Learning for Semantic Video Understanding”. His research is focused on efficient deep learning, object detection, neural attention, spatio-temporal modeling, and language modeling.