Lossy compression of trees via linear directed acyclic graphs

Romain Azaïs (1, 2) , Jean-Baptiste Durand (3) , Christophe Godin (4)

1 BIGS - Biology, genetics and statistics
2 Probabilités et statistiques
3 MISTIS - Modelling and Inference of Complex and Structured Stochastic Systems
4 VIRTUAL PLANTS - Modeling plant morphogenesis at different scales, from genes to phenotype

A classical compression method for trees is to represent them by directed acyclic graphs. This approach exploits subtree repeats in the structure and is efficient only for trees with a high level of redundancy. The class of self-nested trees presents remarkable compression properties because of the systematic repetition of subtrees. In this paper, we provide a better understanding of this specific family of trees and we introduce a lossy compression method that consists in computing the reduction of a self-nested structure that closely approximates the initial data. We compare two versions of our algorithm and a competitive approach of the literature on a simulated dataset.

Mots clés: Unordered trees; Self-nested trees; Directed acyclic graphs; Lossy compression

Domaines: Algorithme et structure de données [cs.DS]

Romain Azaïs, Jean-Baptiste Durand, Christophe Godin. Lossy compression of trees via linear directed acyclic graphs. 2016. ⟨hal-01294013v1⟩