Scientists Harness Machine Learning to Predict Risk of PRRS-PED

Jul 30, 2020
Scientists with the University of Minnesota are harnessing the power of machine learning to make real-time farm-level risk predictions for Porcine Epidemic Diarrhea and Porcine Reproductive and Respiratory Syndrome. Efforts are underway, through the Morrison Swine Health Monitoring Program at the University of Minnesota, to use information collected on a variety of swine diseases to assess the risk of a farm breaking with PED or PRRS.
 
Dr. Paul Sundberg, the Executive Director of the Swine Health Information Center, says the intent is give at risk farms an opportunity to take preventive actions.
 
Clip-Dr. Paul Sundberg-Swine Health Information Center:
 
Participants of the Morrison Swine Health Monitoring Project are reporting their status of their farms and the analysts at the University of Minnesota are taking a look at that information and using a machine learning process that helps to mine that data and look for trends.
 
They not only have disease information about the farms but they also have more characteristic information about the farms as well, the type of farm it is, the topography. They've imported weather data and looked at movements and they're incorporating movement data into the database as well.
 
Through this machine learning process, the one thing that they did was go back to the PED outbreak in 2013 and 2014 and showed how the PED moved from farm to farm and why and the progression and the factors that went into that.
 
The computer is able to then apply that to current conditions and that helps in our predictive models of outbreaks. So we're applying the types of lessons that we learned with PED retrospectively to look proactively out into the future to help us predict outbreaks of PED and we're also working on PRRS as well.
Source : Farmscape
Subscribe to our Newsletters

Trending Video