A Multi-Task Multi-Modal approach to Product Classification
For any retail business, having a consistent understanding of its products is the first step towards being able to understand and forecast customers’ behaviour. In this talk, we discuss the approach that we have taken at ASOS, a global fashion retailer with an ever-evolving catalogue. In particular, we show how Deep Learning can leverage unstructured and partially-labelled data in order to consolidate product attributes.
Fabio is a computer scientist with 8 years of research experience in Evolutionary Computation, Complex Networks and Combinatorial Optimisation. Following his interests in Statistics and Machine Learning, he is currently working as a Data Scientist at ASOS.