Neural Time Series Forecasting & Algorithmic Size Advice at Zalando

Speaker: Kashif Rasul and Julia Lasserre, Zalando

1. Neural Time Series Forecasting
Neural network based forecasting methods have become ubiquitous in large-scale industrial forecasting
applications over the last years. We will providing an introduction and an overview of some of the advances that
have occurred in this area in a hopefully interactive manner.
2. Algorithmic Size Advice at Zalando
Among the many fields where data science is thriving, fashion has the particularity to be worn on a body. This
creates a unique and exciting set of challenges around the topic of size and fit. Zalando has been a pioneer in
tackling the problem and has developed expertise on the fashion/business side of it and on the algorithmic side. In
this talk we will present why size and fit is indeed a difficult problem, what kind of data we use and how we
leverage state-of-the-art learning methods to support various customer segments in their choices. If time allows,
we will dive into the specific case of Size Reco, our size recommender based on customer purchase history.