IDSS conference explores data-disruption in the retail sector

How can retailers extract useful insights from their customer data? How can they translate their data into predictions that help them deliver the right content at the right time? And how can they increase their operational efficiency and even create unique online experiences using machine learning and AI?

Industry and academic leaders explored these questions and more at a day-long conference organized by IDSS called Data-Disruption in the Retail Sector. Speakers included MIT faculty and industry VPs, data scientists, and Chief Algorithms Officers from companies like Stitch Fix, Overstock, Adobe, Salesforce, and Microsoft.

IDSS faculty director Munther Dahleh gave opening remarks on the challenges and opportunities afforded by real-time interactions between people, systems, and data. Statistics and Data Science Center director Devavrat Shah emphasized the scale of the retail industry from manufacturing to supply chain, and from technology to society.

Talk topics included delivering customized experiences using machine learning, optimizing operations with deep learning and natural language processing, and gaining the competitive edge in demand forecasting with artificial intelligence. Speakers shared tips on incorporating technology like RFID and building data science teams to implement new processes.

Prediction and personalization were key themes throughout the day. “The prediction problem in retail,” Shah said, “is demand forecasting for a given product at a given location, time, or channel, and at a given price.”