报告题目：Intelligent Data-Driven Modelling for Marketing
报告人：Bowei Chen，University of Glasgow
报告摘要：Machine learning or in general artificial intelligence has been widely used in real-world marketing applications. Notable examples include Netflix’s movie recommendations and Google’s web search sponsored advertising. In this talk, I will discuss our recent studies which use machine learning algorithms to solve different problems in digital marketing and visual marketing. In the first study, we propose an optimal dynamic model for publishers or sell-side platforms that unifies programmatic guarantee and real-time bidding in display advertising. This study solves the problem of algorithmic pricing and allocation of the non-guaranteed page views into guaranteed contracts. The second study develops a novel computational framework which brings multimedia metrics like contextual relevance, visual saliency and advertisement memorability into real-time bidding. It aims to improve online user's experience as well as maintain the benefits of publisher and advertiser in display advertising. In the third study, we use convolutional neural networks and generative adversarial networks to explore and visualise the fine-grained visual attributes of family cars from their images, examine their effects on sales, and design creative car appearances.
报告人简介：Bowei Chen is an Assistant Professor in Marketing at the University of Glasgow. He received a PhD in Computer Science from the University College London. He works in the cross sections among machine learning, data science and business studies, with a special focus in the applications of probabilistic modelling and deep learning in marketing, finance and information systems.