Active learning to understand infectious disease models and improve policy making.

Modeling plays a major role in policy making, especially for infectious disease interventions but such models can be complex and computationally intensive. A more systematic exploration is needed to gain a thorough systems understanding. We present an active learning approach based on machine learning techniques as iterative surrogate modeling and model-guided experimentation to systematically analyze both common and edge manifestations of complex model runs. Symbolic regression is used for nonlinear response surface modeling with automatic feature selection. First, we illustrate our approach using an individual-based model for influenza vaccination. After optimizing the parameter space, we observe an inverse relationship between vaccination coverage and cumulative attack rate reinforced by herd immunity. Second, we demonstrate the use of surrogate modeling techniques on input-response data from a deterministic dynamic model, which was designed to explore the cost-effectiveness of varicella-zoster virus vaccination. We use symbolic regression to handle high dimensionality and correlated inputs and to identify the most influential variables. Provided insight is used to focus research, reduce dimensionality and decrease decision uncertainty. We conclude that active learning is needed to fully understand complex systems behavior. Surrogate models can be readily explored at no computational expense, and can also be used as emulator to improve rapid policy making in various settings.

Impact of vanilla origins on sensory characteristics of chocolate.

The sensory characteristics of white and milk chocolate with three origins of vanilla (Madagascar, Indonesia, Papua New Guinea) were investigated using a multi-analytical approach. The sensory tests included profiling using Quantitative Descriptive Analysis with a trained panel; Temporal Dominance of Sensations test with untrained respondents from various countries currently residing in Belgium, and the consumers sensory preference test with consumers in Belgium, France, Sweden and Canada. The vanilla extracts were also analyzed by gas chromatography-mass spectrometry (GC-MS) in combination with olfactometry (GC-MS-O) to identify the key aroma compounds in the different vanilla origins. This research shows that sensory profile of Papua New Guinea (PNG) samples is different from Madagascan and Indonesian extracts. The flavor signature was correlated with almondy and anisic descriptors. The perceived difference between Madagascan and Indonesian origins was reduced when assessed within complex matrices such as chocolate. In sugared water, Madagascan vanilla has balsamic, phenolic notes while Indonesian has woodier notes. In a chocolate base, the Indonesian sample lost its woody characteristic for a dairy enhanced characteristic. The botanical variety has more impact on difference in sensory profile of a chocolate than the vanilla origin itself. The sensory data correlates with volatile components identified in this study. Guaiacol and vanillin were identified as the main key aroma compounds in all three vanilla origins using GC-MS-O. Additional key aroma compounds were responsible for the differences in sensory characteristics between the three vanilla origins: anisaldehyde for anisic note in the PNG samples and ethyl 2-methylbutanoate and 3/2-methylbutanoic acid in the Indonesian sample. Sensory profiles of vanillas from various origins were more visible and pronounced in white chocolate compared to milk chocolate. This study gives novel insights to the chocolate, as well as the vanilla sectors, indicating possibilities of vanilla replacement and consumer preferences across Europe and Canada. The consumer study shows that, on average, people do not indicate which chocolate is significantly more liked.