Register Your Forests: Decision Tree Ensemble Optimization by Explicit CPU Register Allocation

arXiv:2404.06846v1 Announce Type: new
Abstract: Bringing high-level machine learning models to efficient and well-suited machine implementations often invokes a bunch of tools, e.g.~code generators, compilers, and optimizers. Along such tool chains, abstractions have to be applied. This leads to not optimally used CPU registers. This is a shortcoming, especially in resource constrained embedded setups. In this work, we present a code generation approach for decision tree ensembles, which produces machine assembly code within a single conversion step directly from the …

abstract abstractions arxiv code compilers cpu cs.lg decision ensemble forests generators leads machine machine learning machine learning models optimization tool tools tree type