Nautilus: An Optimized System for Deep Transfer Learning over Evolving Training Datasets

Summary: Frames DTL adaptation with frozen pre-trained layers as a multi-query optimization problem and introduces two optimizations to share computations across model selections on evolving labeled data. Implemented in Nautilus, it yields up to 5x speedup over naive re-training on benchmark workloads. (summarized by gpt-5-nano on Feb 09 2026)

Paper ID: 6360
Venue: SIGMOD
Year: 2022
Pagerank: 5.5740571e-05
Overall Rank: 7,656 | 47.48%
DOI: 10.1145/3514221.3517846

Incoming Non-self Citations Over Time

Authors

1. Supun Nakandala (University of California San Diego)
2. Arun Kumar (University of California San Diego)

BibTeX Citation

@inproceedings{nakandala_sigmod22,
        title = {{Nautilus: An Optimized System for Deep Transfer Learning over Evolving Training Datasets}},
        author = {Nakandala, Supun and Kumar, Arun},
        series = {{SIGMOD} '22},
        booktitle = {Proceedings of the {ACM} {SIGMOD} International Conference on Management of Data},
        publisher = {Association for Computing Machinery},
        doi = {10.1145/3514221.3517846},
        url = {https://dl.acm.org/doi/10.1145/3514221.3517846},
        year = {2022}
}

Incoming Citations (Sorted by Pagerank)

Showing 5 of 5 citing papers.

Rank	Citing Paper	Year	Venue	Pagerank
6,588	Lotan: Bridging the Gap between GNNs and Scalable Graph Analytics Engines	2023	VLDB	5.833338e-05
9,881	The Image Calculator: 10x Faster Image-AI Inference by Replacing JPEG with Self-designing Storage Format	2024	SIGMOD	5.2040783e-05
10,750	Alsatian: Optimizing Model Search for Deep Transfer Learning	2025	SIGMOD	5.093636e-05
11,189	StarfishDB: a Query Execution Engine for Relational Probabilistic Programming	2024	SIGMOD	5.093636e-05
13,375	Reimagining Deep Learning Systems Through the Lens of Data Systems	2024	VLDB	-

Outgoing Citations (Sorted by Pagerank)

Showing 11 of 11 cited papers.

Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.

Rank	Cited Paper	Year	Venue	Pagerank
205	Snorkel: Rapid Training Data Creation with Weak Supervision	2018	VLDB	0.00025235185
640	Materialization Optimizations for Feature Selection Workloads	2014	SIGMOD	0.00015409494
822	Pipelining in Multi-Query Optimization	2001	PODS	0.00013807229
1,157	Cerebro: A Data System for Optimized Deep Learning Model Selection	2020	VLDB	0.00011924049
1,446	Analyzing and Mitigating Data Stalls in DNN Training	2021	VLDB	0.0001076818
1,569	HELIX: Holistic Optimization for Accelerating Iterative Machine Learning	2019	VLDB	0.00010335423
1,670	MISTIQUE: A System to Store and Query Model Intermediates for Model Diagnosis	2018	SIGMOD	0.00010045615
3,573	VISTA: Optimized System for Declarative Feature Transfer from Deep CNNs at Scale	2020	SIGMOD	7.2977194e-05
4,067	Distributed Deep Learning on Data Systems: A Comparative Analysis of Approaches	2021	VLDB	6.9293511e-05
4,240	LIMA: Fine-grained Lineage Tracing and Reuse in Machine Learning Systems	2021	SIGMOD	6.809685e-05
9,371	Towards an Optimized GROUP BY Abstraction for Large-Scale Machine Learning	2021	VLDB	5.275595e-05

Semantically Similar Papers

#	Overall Rank	Paper	Year	Venue
1	9,962	Structure-Aware Machine Learning over Multi-Relational Databases	2021	SIGMOD
2	4,067	Distributed Deep Learning on Data Systems: A Comparative Analysis of Approaches	2021	VLDB
3	6,708	Serving Deep Learning Models with Deduplication from Relational Databases	2022	VLDB
4	3,541	Where Is My Training Bottleneck? Hidden Trade-Offs in Deep Learning Preprocessing Pipelines	2022	SIGMOD
5	9,737	Saturn: An Optimized Data System for Multi-Large-Model Deep Learning Workloads	2024	VLDB
6	10,750	Alsatian: Optimizing Model Search for Deep Transfer Learning	2025	SIGMOD
7	9,215	Deep Query Optimization	2019	SIGMOD
8	9,414	Model-Parallel Model Selection for Deep Learning Systems	2021	SIGMOD
9	3,764	FastFlow: Accelerating Deep Learning Model Training with Smart Offloading of Input Data Pipeline	2023	VLDB
10	13,375	Reimagining Deep Learning Systems Through the Lens of Data Systems	2024	VLDB