Machine Learning for Precision Public Health (April 16)

The BC Centre for Disease Control (BCCDC) is organizing a special seminar series of Machine Learning for Precision Public Health as part of our large effort to incorporate more data science into our work. These presentations and workshops will cover topics of data-driven decision making, machine learning algorithms, data visualization, and big data ethics. Our next sessions are coming up on Tue April 16, 2019:

BCCDC’s Grand Rounds presentation from 12:00 – 1:00 PM, Hardwick Hall of UBC Medical Student & Alumni Centre (2750 Heather St, Vancouver)

Data, Statistics, and Mathematical Modelling: a Significant Merger, by Dr. Dave Campbell

  • Mechanistic models are built by translating biological and physical dynamics into a mathematical model. Because of their easily interpretable parameters, these models are used for informing actions in outbreaks, describing gene regulatory networks, and exploring the impact of electrical stimulus on muscles. Without data, mathematical science fiction writing is useful for discussion but not prediction or inference. The alternative data rich extreme is to use machine learning to predict the future based on similar past events, but the lack of mechanistic interpretability and biologically plausible structure hinders policy decisions and prevents evidence based mitigating actions. Informing our knowledge by training models with data restricts the model to plausible outcomes consistent with the data and informs evidence based decisions - even in new situations. In this presentation, Dr. Campbell  will provide the intuition behind some of the recent developments in fitting dynamic system models to data, some of the challenges, and some intuition around when to use more or less complex tools.

Hands-on Workshop from 1:15 – 3PM, Room 2264 of Diamond Health Care Centre (2775 Laurel St, Vancouver) *Please note the location change.

A First Glance at Neural Networks for Computer Vision, by Matthew Emery

  • In this introductory workshop, you will learn what neural networks for computer vision are, how you could adapt existing networks made by companies (e.g., Google) to your research, and walk through a hands-on exercise to build a neural network that can classify chest and abdomen X-ray images using R. Please note only a few spaces are still available; registration is required. Participants will need to arrange their own laptop and are recommended to have prior knowledge/experience with R.

For questions or more information, contact Hsiu-Ju Chang ( or Mike Irvine (