||Junhui Wang, City University of Hong Kong
||Xiaojun Mao, School of Data Science, Fudan University
||10:30-11:30, June 24, 2019
||Zibin N201, Fudan University
||Personalized prediction arises as an important yet challenging task, which predicts user-specific preferences on a large number of items given limited information. It is often modeled as certain recommender systems focusing on ordinal or continuous ratings. In this talk, I will present a new collaborative ranking system to predict most-preferred items for each user given search queries. Particularly, a 𝜑-ranker is proposed based on ranking functions incorporating information on users, items, and search queries through latent factor models. Its probabilistic error bound is established showing that its ranking error has a sharp rate of convergence in the general framework of bipartite ranking, even when the dimension of the model parameters diverges with the sample size. Consequently, this result also indicates that the 𝜑-ranker outperforms two major approaches in bipartite ranking: pairwise ranking and scoring. Finally, the proposed 𝜑-ranker is applied to analyze the data from the Mobike big data challenge, consisting of three-million bicycle sharing records.
||Dr. Junhui Wang is Professor in the School of Data Science and Department of Mathematics at City University of Hong Kong. He received his B.S. in Probability and Statistics from Peking University, and Ph.D. in Statistics from University of Minnesota. His research interests include statistical machine learning, unstructured data analysis, model selection and variable selection, as well as their applications in biomedicine, finance and information technology. He has published research articles on leading statistics and machine learning journals, and he also serves as Associate Editor of Annals of the Institute of Statistical Mathematics and Statistics and its interface.