(609d) Adopting Machine Learning As Tool for Delineating the Food-Health Nexus

Lately, consumers around the world are more conscious on the impact of their dietary choices on their health probably more than they have been all through history. This subject has raised significant and sensitive concerns especially on how the choice of ingredients and processing techniques by food manufacturers affect the quality of foods produced. Accordingly, considerable amounts of research and development resources in academia and processing industries has been directed to modeling how each factor of food processing affects the health of the final consumer. Food additives especially sweeteners, preservatives, emulsifiers, and flavor enhancers have been under intense scientific and social scrutiny. However, the delicate need for an objective system of data collection and analyses remains partly satisfied. Such a technique would likely employ data from experimentations conducting using accredited analytical methodologies. Machine learning becomes the optimal selection for these needs. Whether reinforced, supervised, or unsupervised, different machine learning tools such as Principal Component Analysis (PCA), Decision Trees, K-Means Clustering, Generative Mixture Models, Deep Q-Networks (DQNs) and so forth are useful. By way of a simplistic example, the machine learning approach can be used to simplify for an ingredient how increasing its concentration or changing it entirely can affect the blood flow of consumers through several in vivo experimentations. Such studies could even determine how such distinctions in ingredient formulations affect other bodily characteristics such as serum glucose levels of consumers. Machine learning in food formulations can be used to actuate health optimization in reducing undesirable effects of certain products while increasing their goodly benefits. This presentation focuses on providing a comprehensive discourse on how the food and dietary arrangements can be studied and optimized for better health effects on consumers.