HAPEns: Hardware-Aware Post-Hoc Ensembling for Tabular Data
Fuente:
arXiv
Salvato in:
| Autori principali: | , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914386215436288 |
|---|---|
| author | Maier, Jannis Purucker, Lennart |
| author_facet | Maier, Jannis Purucker, Lennart |
| contents | Ensembling is commonly used in machine learning on tabular data to boost predictive performance and robustness, but larger ensembles often lead to increased hardware demand. We introduce HAPEns, a post-hoc ensembling method that explicitly balances accuracy against hardware efficiency. Inspired by multi-objective and quality diversity optimization, HAPEns constructs a diverse set of ensembles along the Pareto front of predictive performance and resource usage. Existing hardware-aware post-hoc ensembling baselines are not available, highlighting the novelty of our approach. Experiments on 83 tabular classification datasets show that HAPEns significantly outperforms baselines, finding superior trade-offs for ensemble performance and deployment cost. Ablation studies also reveal that memory usage is a particularly effective objective metric. Further, we show that even a greedy ensembling algorithm can be significantly improved in this task with a static multi-objective weighting scheme. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_10582 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | HAPEns: Hardware-Aware Post-Hoc Ensembling for Tabular Data Maier, Jannis Purucker, Lennart Machine Learning I.2.6; G.3 Ensembling is commonly used in machine learning on tabular data to boost predictive performance and robustness, but larger ensembles often lead to increased hardware demand. We introduce HAPEns, a post-hoc ensembling method that explicitly balances accuracy against hardware efficiency. Inspired by multi-objective and quality diversity optimization, HAPEns constructs a diverse set of ensembles along the Pareto front of predictive performance and resource usage. Existing hardware-aware post-hoc ensembling baselines are not available, highlighting the novelty of our approach. Experiments on 83 tabular classification datasets show that HAPEns significantly outperforms baselines, finding superior trade-offs for ensemble performance and deployment cost. Ablation studies also reveal that memory usage is a particularly effective objective metric. Further, we show that even a greedy ensembling algorithm can be significantly improved in this task with a static multi-objective weighting scheme. |
| title | HAPEns: Hardware-Aware Post-Hoc Ensembling for Tabular Data |
| topic | Machine Learning I.2.6; G.3 |
| url | https://arxiv.org/abs/2603.10582 |