A novel graph-based multiple kernel learning (GMKL) approach emerged through Hassanzadeh et al's work [11] to accomplish improved kernel-based algorithm performance. A low-rank graph representation enabled the approach to efficiently capture data structures on multiple scales which produced optimal di...
The use of the multivariate normal distribution as a latent construct for modelling observed correlated or associated discrete or ordinal variables can be dated back to the seminal book by Lazarsfeld and Henry (1968) and to the later work by Muthen (1983), in the context of structural equation ...