Our paper was accpeted at ECAI 2025
Our paper, HI-PMK: A Data-Dependent Kernel for Incomplete Heterogeneous Data Representation, was published in the ECAI 2025 proceedings.
The work introduces a data-dependent kernel designed to handle incomplete and heterogeneous data without relying on imputation, supporting mixed numerical and categorical features under different missingness mechanisms (MCAR, MAR, MNAR).