adidas – How product and consumer data can be used for deriving fashion outfits

The sales performance of a fashion product is highly dependent on its ability to be combined with other products in the range to create an outfit. Determining complementary products not only helps to better understand and plan the demand, but also provides a huge opportunity for product recommendations. The presentation will provide an overview on the topic of outfit recommendations in the fashion goods industry and showcase the different steps that needs to be done from a data science perspective to produce meaningful recommendations even in a cold start scenario. In the first part, relevant data sources and preprocessing techniques will be discussed. After that, Machine Learning methods will be presented, that can make use of the data to produce the final recommendations.

Speaker

Dr. René Götz, Senior Data Scientist | adidas

After finishing my master’s degree at the FAU Erlangen/Nuremberg in information systems I started my PhD project on the topic of product recommendations and deriving product similarity using different kinds of data which was in cooperation with adidas. Working now as a Senior Data Scientist I am responsible for different data science capabilities around product performance analytics.