A Maximum Entropy Copula Model for Mixed Data: Representation, Estimation, and Applications
Subhadeep, Mukhopadhyay
备注：
Revised and accepted version. Dedication: This paper is dedicated to E. T. Jaynes, the originator of the Maximum Entropy Principle, for his birth centenary. And to the memory of Leo Goodman, a transformative legend of Categorical Data Analysis. This paper is inspired in part to demonstrate how these two modeling philosophies can be connected and united in some ways
A new nonparametric model of maximum-entropy (MaxEnt) copula density function is proposed, which offers the following advantages: (i) it is valid for mixed random vector. By `mixed' we mean the method works for any combination of discrete or continuous variables in a fully automated manner; (ii) it yields a bonafide density estimate with intepretable parameters. By `bonafide' we mean the estimate guarantees to be a non-negative function, integrates to 1; and (iii) it plays a unifying role in our understanding of a large class of statistical methods. Our approach utilizes modern machinery of nonparametric statistics to represent and approximate log-copula density function via LP-Fourier transform. Several real-data examples are also provided to explore the key theoretical and practical implications of the theory.
Copula-enhanced Vision Transformer for high myopia diagnosis through OU UWF fundus images
Department of Applied Biology；Chemical Technology, The Hong Kong Polytechnic University；Department of Data Science \& AI, The Hong Kong Polytechnic University e2；School of Data Science, Fudan University e1；Department of Applied Mathematics, The Hong Kong Polytechnic University；Department of Biostatistics, City University of Hong Kong；Eye Institute；Department of Ophthalmology, Eye \& ENT Hospital, Fudan University；Eunice Kennedy Shriver National Institute of Child Health；College of Business；Economics, Australian National University
