Technology and predictive analytics are critical for the future of food safety. Determining which predictive analytics techniques work best for a particular challenge is the key to making the most of a predictive analytics solution.
In this episode of our mini video series, Giannis Stoitsis, CTO, and partner at Agroknow, explains how we can find and parameterize the algorithm that will work appropriately for a specific problem and data combination.
Today, there are a plethora of algorithms that can be used. Each algorithm has several parameters that we need to experiment with, in order to get the right prediction model. Then, we need to train and test the model, by applying different splitting methods to our custom dataset, namely a training and testing one.
The easiest way to understand how the algorithms can work their magic is through a real-life example.
An algorithm for food safety is not something to be procured one time off the shelf.
It is developed based on understanding the challenges of the domain of food safety, adapted to meet different constraints and dynamic changes, and evaluated (and re-evaluated) based on its ability to serve goals on a continuing basis.
In order to develop best class predictions, we collaborated with the High-Performance Computing Lab to extend the platform with machine-learning algorithms that have been customized to predict food fraud. Together with the Human-Computer Interaction Research Group of KU Leuven, we investigated new types of dashboards that could represent uncertainty in a prediction, using visual components that humans seem to understand and trust more.
As we have said, choosing the right algorithms is the key to making the most of a predictive analytics solution, but it is just as important to focus on the right prediction metrics to help food safety professionals make insightful decisions.
Stay tuned to find out how to focus on the right prediction metrics in our next episode!
If you’d like to discover how FOODAKAI can help your Food Safety & Quality team prevent product recalls by monitoring & predicting risks, schedule a call with us!
“Funded with the support by European Commission, and more specifically the project CYBELE “FOSTERING PRECISION AGRICULTURE AND LIVESTOCK FARMING THROUGH SECURE ACCESS TO LARGE-SCALE HPC-ENABLED VIRTUAL INDUSTRIAL EXPERIMENTATION ENVIRONMENT EMPOWERING SCALABLE BIG DATA ANALYTICS” (Grant No. 825355)”