Landmark alternating diffusion (Q6983532)

This paper studies an interesting topic, namely that of landmark alternating diffusion. How to understand this idea and how to motivate it? One motivation comes from sensing technology, where in data science it has become more and more common to collect data from multiple sensors to understand the application in mind. Applications of such kinds appear in subjects diverse as statistical methods, machine learning algorithms, Bayesian inference, and mathematical modeling.\N\NAlternating Diffusion (AD) is a commonly used diffusion-based sensor fusion algorithm to deal with the kinds of problems outlined above. The problem with this method though is its high computational complexity.\N\NThe authors propose a variation of AD, called Landmark AD (LAD), which captures the essence of AD while offering superior computational efficiency. They provide a series of theoretical analyses of LAD under the manifold setup and apply it to the automatic sleep stage annotation problem with two electroencephalogram channels to demonstrate its application.