Big data and data-driven modeling: New ways to speed up time-to-market and improve product performance in food and sensory R&D

Session Chair: Ludovic Depoortere, Haystack
Speaker: Katya Vladislavleva, DataStories International
Speaker:  Industry speaker

What is big data and how can we use it for improved predictive power of sensory research?  We will provide some case studies and free access to software/data during the workshop to demonstrate hands on:

  • Data Science for sensory evaluation scientists
  • Data Science for material scientists
  • Data science for formulation scientists
  • Model-guided experimentation for fast data-driven innovation

Effective product & process innovation in consumer research relies on combining domain knowledge and empirical data collected from consumers or trained panels evaluating new products. Interestingly, the vast majority of consumer tests are backward-looking – i.e. they provide an evaluation of the product/process decisions taken by experts and help select optimal configurations(designs).

Predictive analytics and augmented data discovery are modern approaches to innovation that are forward-looking – they recommend what configurations to try to achieve the best expected outcomes.

In this workshop we present an introduction to big data analytics and  hands-on examples of applying predictive analytics techniques for better product design and achieve synergy between historical strength of the consumer research and technologies of today. Several case studies for product optimization in sensory and savoury are presented and one case on brand concept optimization (taste, packaging, messaging).

About Haystack (Ludovic Depoortere)
Ludovic Depoortere is founder of Haystack International NV and has 20 years of experience in sensory & consumer science research worldwide.

About Katya Vladislavleva
Katya Vladislavleva is CEO and chief data scientist at DataStories International NV. Katya’s expertise lies in business-driven data science for food & flavor R&D, and formulation science applications.

« Go Back

Register Now View Programme
Supporting Publication