@article{66209,
  abstract     = {{<jats:p>For timing-sensitive edge applications, the demand for efficient lightweight machine learning solutions has increased recently. Tree ensembles are among the state-of-the-art in many machine learning applications. While single decision trees are comparably small, an ensemble of trees can have a significant memory footprint leading to cache locality issues, which are crucial to performance in terms of execution time. In this work, we analyze memory-locality issues of the two most common realizations of decision trees, i.e., native and if-else trees. We highlight that both realizations demand a more careful memory layout to improve caching behavior and maximize performance. We adopt a probabilistic model of decision tree inference to find the best memory layout for each tree at the application layer. Further, we present an efficient heuristic to take architecture-dependent information into account thereby optimizing the given ensemble for a target computer architecture. Our code-generation framework, which is freely available on an open-source repository, produces optimized code sessions while preserving the structure and accuracy of the trees. With several real-world data sets, we evaluate the elapsed time of various tree realizations on server hardware as well as embedded systems for Intel and ARM processors. Our optimized memory layout achieves a reduction in execution time up to 75 % execution for server-class systems, and up to 70 % for embedded systems, respectively.</jats:p>}},
  author       = {{Chen, Kuan-Hsun and Su, Chiahui and Hakert, Christian and Buschjäger, Sebastian and Lee, Chao-Lin and Lee, Jenq-Kuen and Morik, Katharina and Chen, Jian-Jia}},
  issn         = {{1539-9087}},
  journal      = {{ACM Transactions on Embedded Computing Systems}},
  number       = {{6}},
  pages        = {{1--26}},
  publisher    = {{Association for Computing Machinery (ACM)}},
  title        = {{{Efficient Realization of Decision Trees for Real-Time Inference}}},
  doi          = {{10.1145/3508019}},
  volume       = {{21}},
  year         = {{2022}},
}