 In this episode, we will explain how one can find and parameterize the algorithm that will work best for the specific problem and data combination. Given that I have access to historical data about all food product recalls and border rejections that have taken place until December 2019, classified according to the raw materials, ingredients and hazards that were linked to the recall. Is it possible to predict with high confidence how many food safety incidents we are going to have for each ingredient category in 2020? Well, by extensive testing and parameterization, this means trying as many algorithms and variations as we can. We worked with four different families of supervised machine learning algorithms and one family of deep learning algorithms. We had to try different parameters for each algorithm, execute different paths over the input data to see what type of prediction models were built, refine and revise and rerun again. Surprisingly, the algorithm that performed better has not been the one that we originally expected. We therefore ran two complete iterations, one producing predictions based on the global food incidents dataset and one using the data specifically about chocolate products and ingredients. What kind of results did we get for critical ingredients used in chocolate products? We took the peanuts as an example and the algorithm is over 98% confident that in 2020, aflatoxin incidents will continue. What I'm saying is that 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 best prediction model. That's all for today. Thank you for watching. Don't forget to subscribe to our weekly Food Safety Intelligence newsletter to stay up to date with the latest industry news. Find the link below the video.