mlwhatif: What If You Could Stop Re-Implementing Your Machine Learning Pipeline Analyses Over and Over?

Summary: mlwhatif declaratively specifies data-centric what-if analyses over existing ML pipelines, automatically generating and optimizing pipeline variants. It supports robustness to data errors, cleaning impact, and preprocessing effects on fairness without reimplementation. (summarized by gpt-5.6-luna on Jul 24 2026)

Paper ID: hdb2a4a516805d97b
Venue: VLDB
Year: 2023
Pagerank: 5.4596612e-05
Overall Rank: 7,651 | 48.72%
DOI: 10.14778/3611540.3611606
PDF: Download (CC BY-NC-ND 4.0)

Incoming Non-self Citations Over Time

Authors

1. Stefan Grafberger (University of Amsterdam)
2. Shubha Guha (University of Amsterdam)
3. Paul Groth (University of Amsterdam)
4. Sebastian Schelter (University of Amsterdam)

BibTeX Citation

@article{grafberger_vldb23,
        title = {{mlwhatif: What If You Could Stop Re-Implementing Your Machine Learning Pipeline Analyses Over and Over?}},
        author = {Grafberger, Stefan and Guha, Shubha and Groth, Paul and Schelter, Sebastian},
        journal = {PVLDB},
        series = {{VLDB} '23},
        volume = {16},
        number = {12},
        pages = {4002--4005},
        doi = {10.14778/3611540.3611606},
        url = {https://doi.org/10.14778/3611540.3611606},
        year = {2023}
}

Incoming Citations (Sorted by Pagerank)

Showing 3 of 3 citing papers.

Rank	Citing Paper	Year	Venue	Pagerank
8,229	Falcon: Fair Active Learning using Multi-armed Bandits	2024	VLDB	5.3578737e-05
9,759	Shapley Value Estimation Based on Differential Matrix	2025	SIGMOD	5.115551e-05
10,492	ASSS: Adaptive Stratified Sampling for Shapley-like Values	2026	SIGMOD	4.9605343e-05

Outgoing Citations (Sorted by Pagerank)

Showing 5 of 5 cited papers.

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

Rank	Cited Paper	Year	Venue	Pagerank
897	Efficient Task-Specific Data Valuation for Nearest Neighbor Algorithms	2019	VLDB	0.00013162662
1,163	Responsible Data Management	2020	VLDB	0.00011709266
4,628	MLINSPECT: A Data Distribution Debugger for Machine Learning Pipelines	2021	SIGMOD	6.4796471e-05
7,558	Automating and Optimizing Data-Centric What-If Analyses on Native Machine Learning Pipelines	2023	SIGMOD	5.4788791e-05
11,868	Screening Native ML Pipelines with “ArgusEyes”	2022	CIDR	4.9605343e-05

Semantically Similar Papers

#	Overall Rank	Paper	Year	Venue
1	9,454	Towards Observability for Production Machine Learning Pipelines	2022	VLDB
2	6,234	Materialization and Reuse Optimizations for Production Data Science Pipelines	2022	SIGMOD
3	10,925	PipeLens: Identifying Interventions for Resolving Malfunctioning Data Science Pipelines	2026	VLDB
4	5,466	HAIPipe: Combining Human-generated and Machine-generated Pipelines for Data Preparation	2023	SIGMOD
5	11,466	mlidea: Interactively Improving ML Data Preparation Code via “Shadow Pipelines”	2025	VLDB
6	2,116	Production Machine Learning Pipelines: Empirical Analysis and Optimization Opportunities	2021	SIGMOD
7	11,723	Reconstructing and Querying ML Pipeline Intermediates	2023	CIDR
8	6,338	Lightweight Inspection of Data Preprocessing in Native Machine Learning Pipelines	2021	CIDR
9	4,628	MLINSPECT: A Data Distribution Debugger for Machine Learning Pipelines	2021	SIGMOD
10	7,558	Automating and Optimizing Data-Centric What-If Analyses on Native Machine Learning Pipelines	2023	SIGMOD