Topic：Machine learning based customer behavior analysis for retail industry

Presenter:Yuzhong Chen

Time:May 7,10:30am

Location:Room 5004,Physsics building

Abstract:

Machine learning/deep learning based technologies have been widely used for marketing purposes in retail industry, such as customer behavior analysis, product recommendation, supply chain prediction, and promotion optimization. This talk mainly covers two topics: (1) the implementation of a consumer evaluation system based on Random Forest, Logistic Regression with Elastic-Net, NMF, PCA, and Conditional Neural Chain for predicting customer purchasing behavior in order to deliver highly personalized marketing messages to individuals; (2) Recurrent Neural Network based customer behavior analysis algorithm of Kroger’s early warning system for churning prediction.

CV:

I’m currently a machine learning research scientist in the Science team of 84.51 LLC, a data analysis company of Kroger. My research work mainly focuses on developing neural network based approaches to solve prediction and optimization problems in retail industry. Previously, I worked as a data scientist focusing on implementing production level code for customer evaluation engine based on various machine learning techniques in Conversant LLC. Before joining the industry, I was awarded a Ph.D. degree in Electrical Engineering in Arizona State University in May 2016, under the supervision of Prof. Ying-Cheng Lai, and my Ph.D. thesis was Predicting and controlling complex networks.