Strange Loop

From Video Games to Fashion: a Machine Learning Journey

Gaming is a rapidly growing industry. Players generate extremely rich datasets that record each and every in-game action, social interaction, ad view or purchase. Because of the constant and continued relationship between player and game, player behavioral data constitutes a unique source of information on human behavior: a simplified dynamic system that replicates real life. 

Fashion goes beyond a basic need: it is a way to express our own individuality. The fashion industry is one of the largest in the world, representing 2 percent of global GDP, and has strong social and economic implications worldwide. Vast amounts of data are constantly generated, ranging from customer behavior to complex supply-chain systems. Success in the fashion industry relies on carefully balancing art and science.

In this talk I will discuss the common challenges to both industries and the emergency of models that scale to really big datasets and manage to capture various kinds of consumer tastes and behaviors. Because of the importance of the temporal dimension, models based on CNN, LSTM or DeepAR play a crucial role to push back this new technological frontier. I will review how these techniques can be used to create customized events and provide personalized predictions.

Africa Perianez

Africa Perianez


África Periáñez is the Chief Analytics Officer at Inditex. África leads the research, application and strategic development of data science across the organization. Her goal is to build robust machine learning products to champion fashion retailing experience. Before landing in fashion industry África founded Yokozuna Data, an AI company headquartered in Tokyo that built a pioneer machine learning platform to predict videogame player behavior. África holds a PhD in Mathematics from the University of Reading, and MSc's in String Theory and Theoretical Physics from CERN and the Autonomous University of Madrid. África has been Marie Curie EU research fellow at CERN and scientist at RIKEN, Japan (using K computer) and the German Weather Service (working in satellite data assimilation). She is co-author of multiple peer-reviewed articles and regular invited speaker in international conferences. África has solid industry experience at SPSS Inc., IBM and Silicon Studio. Her work has been featured in media such as Bloomberg, Quartz, Nikkei, El País, SINC, The Economic Times and The Japan Times.