Abstract: Every day we interact with machine learning systems that personalize their predictions to individual users, whether to recommend movies, find new friends or dating partners, or organize our news feeds. Such systems involve several modalities of data, ranging from sequences of clicks or purchases, to rich modalities involving text, images, or social interactions.
In this talk we'll introduce a common set of principles and methods that underpin the design of personalized predictive models. We'll begin by revising "traditional" forms of personalized learning, such as recommender systems. Later, we'll see how similar ideas apply to domains such as natural language processing and computer vision. Finally, we'll study the consequences and risks of deploying personalized predictive systems.
Bio: Julian McAuley has been a professor in the Computer Science Department at the University of California, San Diego since 2014. Previously he was a postdoctoral scholar at Stanford University after receiving his PhD from the Australian National University in 2011. His research is concerned with developing predictive models of human behavior using large volumes of online activity data.