# Ensembles at Any Cost? Accuracy-Energy Trade-offs in Recommender Systems

arXiv: https://arxiv.org/abs/2604.07869 · HTML: https://mlsanity.com/paper/2604.07869/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 17 of its 31 references that are in the map.
- First posted on arXiv 9 Apr 2026.

## Builds on (most cited first, 17 of 17)

- [Green AI](https://mlsanity.com/paper/1907.10597/) (2019)
- [Towards the Systematic Reporting of the Energy and Carbon Footprints of Machine Learning](https://mlsanity.com/paper/2002.05651/) (2020)
- [Towards Green Automated Machine Learning: Status Quo and Future Directions](https://mlsanity.com/paper/2111.05850/) (2021)
- [LensKit for Python: Next-Generation Software for Recommender System Experiments](https://mlsanity.com/paper/1809.03125/) (2018)
- [From Clicks to Carbon: The Environmental Toll of Recommender Systems](https://mlsanity.com/paper/2408.08203/) (2024)
- [Assembled-OpenML: Creating Efficient Benchmarks for Ensembles in AutoML with OpenML](https://mlsanity.com/paper/2307.00285/) (2023)
- [Presentation of a Recommender System with Ensemble Learning and Graph Embedding: A Case on MovieLens](https://mlsanity.com/paper/2008.01192/) (2020)
- [CMA-ES for Post Hoc Ensembling in AutoML: A Great Success and Salvageable Failure](https://mlsanity.com/paper/2307.00286/) (2023)
- [Q(D)O-ES: Population-based Quality (Diversity) Optimisation for Post Hoc Ensemble Selection in AutoML](https://mlsanity.com/paper/2307.08364/) (2023)
- [Green Recommender Systems: Optimizing Dataset Size for Energy-Efficient Algorithm Performance](https://mlsanity.com/paper/2410.09359/) (2024)
- [The Potential of AutoML for Recommender Systems](https://mlsanity.com/paper/2402.04453/) (2024)
- [EMERS: Energy Meter for Recommender Systems](https://mlsanity.com/paper/2409.15060/) (2024)
- [e-Fold Cross-Validation for Recommender-System Evaluation](https://mlsanity.com/paper/2412.01011/) (2024)
- [Green Recommender Systems: Understanding and Minimizing the Carbon Footprint of AI-Powered Personalization](https://mlsanity.com/paper/2509.13001/) (2025)
- [Eco-Aware Graph Neural Networks for Sustainable Recommendations](https://mlsanity.com/paper/2410.09514/) (2024)
- [Algorithm Selection for Recommender Systems via Meta-Learning on Algorithm Characteristics](https://mlsanity.com/paper/2508.04419/) (2025)
- [The Environmental Impact of Ensemble Techniques in Recommender Systems](https://mlsanity.com/paper/2511.11649/) (2025)

## Abstract (arXiv)

Ensemble methods are frequently used in recommender systems to improve accuracy by combining multiple models. Recent work reports sizable performance gains, but most studies still optimize primarily for accuracy and robustness rather than for energy efficiency. This paper measures accuracy energy trade offs of ensemble techniques relative to strong single models. We run 93 controlled experiments in two pipelines: 1. explicit rating prediction with Surprise (RMSE) and 2. implicit feedback ranking with LensKit (NDCG@10). We evaluate four datasets ranging from 100,000 to 7.8 million interactions (MovieLens 100K, MovieLens 1M, ModCloth, Anime). We compare four ensemble strategies (Average, Weighted, Stacking or Rank Fusion, Top Performers) against baselines and optimized single models. Whole system energy is measured with EMERS using a smart plug and converted to CO2 equivalents. Across settings, ensembles improve accuracy by 0.3% to 5.7% while increasing energy by 19% to 2,549%. On MovieLens 1M, a Top Performers ensemble improves RMSE by 0.96% at an 18.8% energy overhead over SVD++. On MovieLens 100K, an averaging ensemble improves NDCG@10 by 5.7% with 103% additional energy. On Anime, a Surprise Top Performers ensemble improves RMSE by 1.2% but consumes 2,005% more energy (0.21 vs. 0.01 Wh), increasing emissions from 2.6 to 53.8 mg CO2 equivalents, and LensKit ensembles fail due to memory limits. Overall, selective ensembles are more energy efficient than exhaustive averaging,