Are encoders able to learn landmarkers for warm-starting of Hyperparameter Optimization?

Antoni Zajko; Katarzyna Wo\'znica

REVIEW 4 major objections 6 minor 23 references

Are encoders able to learn landmarkers for warm-starting of Hyperparameter Optimization?

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper asks whether neural encoders can learn tabular dataset representations that capture landmarker information and whether that helps warm-start Bayesian hyperparameter optimization (HPO).

desk verdict Useful but not conclusive: the paper's 'alignment doesn't transfer' claim needs an oracle warm-start baseline before it holds. read the letter →

arxiv 2507.12604 v1 pith:IAWB5UU5 submitted 2025-07-16 cs.LG

Antoni Zajko , Katarzyna Woźnica This is my paper · ORCID

classification cs.LG

keywords tabulardatasetrepresentationlandmarkerswarm-startingBayesianhyperparameteroptimizationmetriclearningDataset2Vecmeta-learning

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper asks whether neural encoders can learn tabular dataset representations that capture landmarker information and whether that helps warm-start Bayesian hyperparameter optimization (HPO). It proposes two encoders, one trained by metric learning to align representation distances with landmarker distances and one trained to reconstruct landmarker vectors. The encoders do learn landmarker-aligned representations, with distance correlations rising well above the Dataset2Vec baseline. However, in warm-starting Bayesian HPO, neither encoder yields a statistically significant improvement over random or heuristic portfolio baselines after the full optimization run. The paper concludes that distance alignment with landmarkers may be necessary but is not sufficient for this meta-task.

What carries the argument

The load-bearing mechanism is the Requirement formulated in the introduction: distances between representations of datasets should align with distances between the corresponding landmarkers, where a landmarker vector records the performances of a fixed portfolio of hyperparameter configurations on a dataset. The paper implements this requirement in two loss functions. The metric-learning loss (Eq. 1) minimizes the mean squared difference between the Euclidean distance $d(\varphi(X^{(i,1)}, y^{(i,1)}), \varphi(X^{(i,2)}, y^{(i,2)}))$ of two encoded datasets and the Euclidean distance $d(l^{(i,1)}, l^{(i,2)})$ of their landmarker vectors. The reconstruction loss (Eq. 2) minimizes the mean squared error between a predicted landmarker vector and the true one, using the predicted vector as the dataset representation. The evaluation metric (Eq. 3) measures the Spearman correlation between representation distances and landmarker distances, and it is this correlation that the paper uses to judge whether the Requirement is fulfilled.

What would settle it

Run the same warm-start protocol but replace encoder representations with the exact landmarker vectors in the nearest-neighbor search; if this oracle representation fails to improve over random portfolio selection at the end of optimization, the paper's claim that HPO is largely robust to warm-start selection is confirmed, while if it succeeds, the encoders' weak alignment is the bottleneck.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is a negative result with a positive component: the proposed encoders can satisfy the proposed Requirement, that Euclidean distances between dataset representations align with Euclidean distances between landmarker vectors, but this alignment does not translate into significant gains in warm-starting Bayesian HPO. Training Dataset2Vec with a metric-learning loss raises the Spearman correlation between representation distances and landmarker distances from 0.037 to 0.332, and the reconstruction approach reaches 0.470 when distances are measured to predicted landmarkers. Yet in the warm-start evaluation, after the full Bayesian HPO run, even the true-landmarker selection method is not statistically distinguishable from no warm-start or random portfolio selection, and the encoder-based methods are statistically indistinguishable from other portfolio-based baselines. The paper thus establishes that the link between representation-to-landmarker distance alignment and downstream HPO performance is weak.

Load-bearing premise

The paper assumes that making dataset-representation distances match landmarker distances is enough signal to produce useful warm-start configurations for Bayesian HPO; if that assumption fails, the whole encoder-design strategy loses its justification.

Editorial extensions

If this is right

Evaluating encoders by how well their distances correlate with landmarker distances can be misleading, because high correlation does not imply usefulness in warm-starting HPO.
Encoder-based warm-starting can partially substitute true landmarkers in the initial phase of Bayesian HPO, as the proposed methods are statistically indistinguishable from the landmarker upper bound just after warm-start.
After a full Bayesian HPO run, the choice of warm-start approach has no statistically significant effect, since even the true-landmarker method is indistinguishable from random portfolio selection.
General-purpose representations like plain Dataset2Vec carry almost no landmarker information, with a correlation of 0.037, so task-specific training is needed if landmarker information is the goal.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A natural extension the authors leave implicit is to test the distance-alignment requirement on other meta-tasks, such as model selection or dataset recommendation, where neighbor quality may matter more than in a 20-iteration Bayesian HPO run.
The results suggest that portfolio construction itself, with 100 configurations chosen by tournament selection over clusters, may dominate warm-start quality; representation learning might show larger gains with a larger or more carefully stratified portfolio.
One could design a stricter test of the Requirement by training the encoder to preserve only the ordering of landmarker distances to the target dataset, which is the information the nearest-neighbor warm-start step actually uses.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Quick take: this is a genuinely useful negative result—training encoders to align with landmarkers doesn't automatically buy you HPO warm-start gains—but the paper's central claim is not as solid as the abstract implies, because the landmarker baseline they treat as an upper bound is only a k-NN selector and may be a weak one.

What's good: the paper does a clean job of setting up two encoder objectives (metric learning on landmarker distances, landmarker reconstruction), evaluating them against Dataset2Vec, and including sensible baselines (random from portfolio, rank heuristic). The honest discussion in Section 5 is a plus. The finding that representations can correlate with landmarkers (0.47 for reconstruction, 0.33 for metric) yet still not improve warm-starting after full optimization is worth reporting, and the CD plots make the null result clear. The protocol itself is reusable.

Where it's soft: the stress-test note is right. The 'Landmarkers' baseline in Section 4.4 is described as an upper bound, but it's just nearest neighbors in landmarker space with one configuration pulled from each neighbor. There's no reason that's the best possible warm-start, and there's no oracle test (e.g., take the portfolio configuration that actually performs best on the target dataset). Without that, saying 'even landmarkers don't help' is really saying 'this particular k-NN warm-start algorithm doesn't help.' The authors list the algorithm as a possible cause in Section 5, but they don't isolate it. That's the main gap.

Also: the reconstruction loss in Section 4.5.1 (0.0307) is reported as lower than the initial value (0.0256), which is backwards; that looks like a typo but undermines the 'all approaches learn' claim. The distance-correlation metric (Eq. 3, Table 1) is somewhat self-referential for the metric encoder, since Eq. 1 optimizes exactly that; it's a legitimate held-out generalization check, but it shouldn't be read as independent evidence of usefulness. Reproducibility is partial: anonymous URL without versioning, and the OpenML dataset IDs aren't listed.

Bottom line: this is for AutoML/meta-learning folks. It deserves a serious referee, but not as-is. The main fix is to add an oracle or stronger landmarker-based warm-start to test whether the negative result is about the encoders or the selection mechanism. Then it would be a solid empirical paper.

Referee Report

4 major / 6 minor

Summary. The paper proposes two encoder-based methods for representing tabular datasets, specifically designed for warm-starting Bayesian hyperparameter optimization (HPO). The core idea is a "Requirement": distances between dataset representations should align with distances between landmarker vectors (performance vectors of a portfolio of configurations). The first method trains a Dataset2Vec-style encoder with a metric-learning loss (Eq. 1) that minimizes the MSE between representation distances and landmarker distances. The second method adds a reconstruction head that predicts the landmarker vector from the latent representation, using the predicted vector as the dataset representation (Eq. 2). The encoders are evaluated on a meta-dataset of ~700 OpenML binary classification tasks with xgboost, using (i) distance correlation (Eqs. 3–4, Table 1) and (ii) warm-start gains in Bayesian HPO measured by ADTM and CD plots (Figs. 3–4). The main findings are that the proposed encoders achieve higher correlation with landmarker distances than the Dataset2Vec baseline, but these representations do not yield statistically significant improvements over random or heuristic baselines after full HPO optimization; even the true-landmarker baseline is not statistically distinguishable from baselines after 20 iterations. The authors conclude that encoder alignment with landmarkers may not directly translate to HPO warm-starting gains.

Significance. If the results are taken at face value, the paper provides a valuable negative result for the meta-learning community: it demonstrates that optimizing a dataset-representation objective to match landmarker distances does not automatically improve downstream HPO warm-starting. This challenges the assumption that such distance alignment is a sufficient signal for transfer in Bayesian HPO. The paper also contributes two concrete encoder architectures and a reproducible evaluation pipeline, with code provided. The evaluation design is thoughtful in separating the alignment metric from the target meta-task. However, the central negative claim is currently confounded by the warm-start selection algorithm used for all methods, including the landmarker baseline, which prevents the paper from establishing that the representation objective itself is the limiting factor. If the oracle experiment recommended below confirms the results, the paper would be a solid, useful contribution.

major comments (4)

[§4.4 and §4.5.3 (Landmarkers baseline and Warm-start results)] The landmarker baseline is described as an upper bound (§4.4, item 4), but it is implemented as a k-NN selector that retrieves the k=5 nearest datasets in landmarker space and proposes the best configuration from each. This is not an upper bound for warm-starting: Euclidean nearest neighbors in a 100-dimensional landmarker space need not be the datasets whose best configurations transfer to the target, and choosing one configuration per neighbor is a weak way to use even perfect landmarker information. Consequently, the observation in §4.5.3 and Figure 4b that the landmarker baseline is statistically indistinguishable from random after 20 HPO iterations may reflect the selection algorithm rather than the informative content of landmarkers or the proposed encoders. To support the abstract's claim that encoder alignment with landmarkers 'may not directly translate' to HPO gains, the authors should add an oracle baseline: for each target dataset, select the portfolio configuration that has the best true landmarker performance (or the best actual performance after evaluation) as the warm-start point. If that oracle also shows no significant gains after full optimization, then the negative conclusion would be sound. Until such an experiment is run, the central claim is not fully established.
[§4.5.1 (Encoders training)] The numerical comparison for the landmark reconstructor is internally inconsistent: the text states 'landmarker reconstructor with reconstruction loss of 0.0307, which is lower than the initial value at the beginning of the training – 0.0256'. Since 0.0307 > 0.0256, the claim that the loss decreased is false under the stated numbers. Either the reported final loss should be lower than the initial loss (e.g., swapping the values or correcting the printed figures), or the sentence should be rewritten to say the final loss is higher, which would contradict the claim that the method learned. This is a load-bearing detail because it is part of the evidence that the reconstructor successfully minimized its objective. Please correct the numbers and the comparison.
[Eq. (3) and Table 1] The distance-correlation metric defined in Eq. (3) measures exactly the alignment that the metric-learning encoder's loss (Eq. (1)) optimizes during training, namely the correspondence between Euclidean distances of representations and Euclidean distances of landmarkers. Therefore, the high correlation reported for the metric-based encoder (0.332) in Table 1 is, to a substantial degree, a check of whether the training objective generalizes to held-out dataset pairs, rather than an independent confirmation that the representation captures landmarker properties beyond the optimized objective. The paper should explicitly acknowledge this overlap and, ideally, add an evaluation that is not directly aligned with the training loss—for example, the correlation of the predicted landmarker distances (Eq. 4) for the metric method, or a downstream task that tests transfer of the representation without being part of the loss.
[Table 1 and §4.3.1] Table 1 mixes two different correlation formulas without clear labeling: for Dataset2Vec basic and metric learning, it reports Eq. (3) (correlation between representation distances and landmarker distances), whereas for the reconstruction approach it reports Eq. (4) (correlation between distances from reconstructed landmarkers to true landmarkers and true landmarker distances). These metrics are not directly comparable, yet the table presents them side by side under a single 'Correlation' column. This makes the cross-method comparisons in the table misleading. The authors should either report a single common metric for all methods (e.g., Eq. (3) for all, or a separate table column for each metric) or clearly annotate which rows use which equation and avoid implicit comparisons across rows.

minor comments (6)

[§4.2] The sentence 'The is performed in three steps' contains a typo; it should read 'This is performed in three steps' or 'The selection is performed in three steps'.
[§4.5.3] The reference to 'In Figure 3a' and 'In Figure 3b' is ambiguous: the text says 'In Figure 3a, we present an ADTM plot... In Figure 3b, we extend that comparison showing scaled values...' but it is not clear which panel corresponds to which description. Also, the sentence 'the landmarker reconstruction is better than metric learning' should cite the specific panel and metric supporting this claim.
[Eq. (1) and Eq. (2)] The notation for indices is unclear: in Eq. (1), the pair index i is used to sum over N pairs, but the terms inside the summation are written with superscripts (i,1) and (i,2). Please clarify the indexing of the datasets and the summation bounds. Similarly, in Eq. (2), the subscript n is used in the sum but the terms are written with index i; the notation should be made consistent.
[Table 1] The standard deviations are reported over S=20 samples, but the number of dataset pairs N=1000 is not stated in the table caption. Please include the sample size and the definition of 'correlation' (Spearman) directly in the caption, and add a statistical test (e.g., confidence intervals or p-values) for the claim that correlations are 'significantly higher than 0'.
[§1 and §5] The term 'formulated by ourselves' in the abstract is informal; consider replacing with 'proposed'. Also, in §5, the phrase 'too weak information' is grammatically incorrect; it should be 'too weak' or 'insufficient information'.
[§2] 'meta-validation serves as the set of data on which we calculate metrics' should be 'set of datasets' (or 'data') for grammatical agreement. Additionally, the dashed/dotted/solid line explanation in Figure 1 is not referenced in the text; please add a brief description of what each line type indicates.

Circularity Check

1 steps flagged · score 4.0 of 10

Distance-correlation evaluation is a self-referential check of the training objective, but the central HPO warm-start claim is independently tested.

self definitional [Section 3.2 (Eq. 1), Section 4.3.1 (Eq. 3), Table 1, Abstract]
"Distances between representations of the datasets should align with the distances between corresponding landmarkers. ... 𝑙𝑜𝑠𝑠 𝑚𝑒𝑡𝑟𝑖𝑐 = 1/𝑁 ∑ 𝑑(𝜑(𝑋(𝑖,1), 𝑦(𝑖,1)), 𝜑(𝑋(𝑖,2), 𝑦(𝑖,2)))^2 − 𝑑(𝑙(𝑖,1), 𝑙(𝑖,2))^2 (1) ... 𝑐𝑜𝑟𝑟([𝑑(𝜑(𝑋(𝑖,1), 𝑦(𝑖,1)), 𝜑(𝑋(𝑖,2), 𝑦(𝑖,2)))] 𝑖=1,...,𝑁 ,[𝑑(𝑙(𝑖,1), 𝑙(𝑖,2))] 𝑖=1,...,𝑁) (3)"

The metric-learning encoder is trained (Eq. 1) to minimize the discrepancy between the distance vectors d(φ,φ) and d(l,l); the primary evaluation metric (Eq. 3) is the Spearman correlation between exactly those same two distance vectors, reported in Table 1 as evidence that the encoders 'can effectively learn representations aligned with landmarkers' (Abstract). On training pairs, low Eq. 1 loss forces this correlation near 1 by construction; on the held-out pairs in Table 1 it remains a generalization check of the same objective, not an independent meta-task measure. For the reconstruction variant the relation is even more direct, since the representation is the predicted landmarker vector. The HPO warm-start evaluation (Figs.

full rationale

The core derivation chain — define the landmarker-alignment Requirement, train encoders with Eq. 1/Eq. 2, and then test warm-start gains in Bayesian HPO (Figs. 3-4) — is self-contained and does not reduce to its inputs. The negative HPO finding is the paper's main empirical claim and is evaluated against independent baselines (no warm-start, random from portfolio, rank heuristic) and the true-landmarker selector. The distance-correlation metric (Eq. 3, Table 1) is a self-referential diagnostic: it measures the same alignment that the metric loss (Eq. 1) enforces, so the abstract's statement that encoders 'can effectively learn representations aligned with landmarkers' is partly a restatement of training-objective generalization; the paper transparently labels this metric as 'the degree of fulfillment of the proposed requirement.' There is one self-citation (Płudowski et al. 2024, sharing two present authors) used to motivate task-specific representations, but the paper independently tests the general Dataset2Vec baseline (Table 1, corr 0.037) and reproduces the insufficiency, so the citation is not load-bearing. The landmarker baseline is called an 'upper bound' but is only a k-NN selector; this threatens the validity of the negative conclusion (the true-landmarker baseline may itself be weak), but that is an experimental-design concern, not circularity. No uniqueness theorems or ansatz-smuggling citations appear.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The ledger is light because the paper proposes no new entities or physical/mathematical axioms. The free parameters are evaluation hyperparameters, and the main domain assumptions are the landmarker-distance proxy and the sufficiency of Dataset2Vec's architecture. The 'Requirement' is a self-imposed objective rather than an independent empirical law.

free parameters (4)

portfolio size (number of clusters) = 100
Number of xgboost configurations selected via k-means clustering and tournament selection. Hand-chosen, not fitted to target data.
warm-start count = 5
Number of initial configurations given to Bayesian HPO. Hand-chosen.
HPO iterations = 20
Number of Bayesian HPO iterations. Hand-chosen.
correlation sample pairs S = 20
Number of 1000-pair samples used to average Spearman correlations in Eq. 3 and Eq. 4. A computational shortcut chosen by the authors.

assumptions (3)

domain assumption Landmarker distance is a useful proxy for dataset similarity for warm-starting
Section 1 defines the 'Requirement' that aligning representation distances with landmarker distances improves warm-starting; the paper only tests this indirectly via downstream HPO.
domain assumption Dataset2Vec (DeepSets) has the capacity to encode the necessary dataset structure
Both proposed encoders build on Dataset2Vec from Jomaa et al. 2021 without reassessing its expressivity.
domain assumption Euclidean distance is the appropriate metric in both representation and landmarker space
Eqs. 1, 3, and 4 all use Euclidean distance without ablation or justification.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Are encoders able to learn landmarkers for warm-starting of Hyperparameter Optimization?." pith.science (2026). https://pith.science/paper/IAWB5UU5

@misc{pith2026250712604,
  author       = {Pith},
  title        = {Pith review of: Are encoders able to learn landmarkers for warm-starting of Hyperparameter Optimization?},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IAWB5UU5}},
  note         = {Machine review of arXiv:2507.12604}
}

read the original abstract

Effectively representing heterogeneous tabular datasets for meta-learning purposes is still an open problem. Previous approaches rely on representations that are intended to be universal. This paper proposes two novel methods for tabular representation learning tailored to a specific meta-task - warm-starting Bayesian Hyperparameter Optimization. Both follow the specific requirement formulated by ourselves that enforces representations to capture the properties of landmarkers. The first approach involves deep metric learning, while the second one is based on landmarkers reconstruction. We evaluate the proposed encoders in two ways. Next to the gain in the target meta-task, we also use the degree of fulfillment of the proposed requirement as the evaluation metric. Experiments demonstrate that while the proposed encoders can effectively learn representations aligned with landmarkers, they may not directly translate to significant performance gains in the meta-task of HPO warm-starting.

Figures

Figures reproduced from arXiv: 2507.12604 by the authors.

**Figure 1.** Visualization of the relation between data, encoders, and results. The dashed lines correspond to datasets, the dotted line corresponds to intermediate results, and the solid line depicts what is evaluated in the proposed methods. serves as the set of data on which we calculate metrics to assess the quality of the encoders. The detailed data preparation procedure description is in Section 4.1. Having meta-training a… view at source ↗

**Figure 2.** Schematic view of the Equation 1, which calculates the Mean Squared distance between Euclidean distances of meta-features and landmarkers. The first component (top) is the distance between feature representations, and the second (bottom) is the distance between landmarkers. 3.3 Representations based on the landmarkers reconstruction The second method is based on a different way of satisfying the Requirement. It crea… view at source ↗

**Figure 3.** Comparison of performance between warm-start approaches. The dotted line indicates the moment when warm-start ends and optimization begins. too weak information about the landmarkers that are yielded by encoders. The second cause is the fact that the Bayesian HPO does not benefit from the specific warm-start algorithm but the set of configurations from the specific hyperparameter subspace. On the other hand, we show… view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: Comparison of warm-start approaches on CD plots. The more the vertical bar is to the right, the better. Approaches connected with the horizontal bar are statistically indistinguishable regarding the Wilcoxon-Holm test. Social impact statement After careful reflection, …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

23 extracted references · 10 canonical work pages

[1]

A. Alanazi. Using machine learning for healthcare challenges and opportunities. Informatics in Medicine Unlocked, 30: 0 100924, 2022. doi:10.1016/j.imu.2022.100924

arXiv 2022
[2]

T. Bai, Y. Li, Y. Shen, X. Zhang, W. Zhang, and B. Cui. Transfer learning for bayesian optimization: A survey, 2023. URL https://arxiv.org/abs/2302.05927

arXiv 2023
[3]

Bentéjac, A

C. Bentéjac, A. Csörgő, and G. Martínez-Muñoz. A comparative analysis of gradient boosting algorithms. Artificial Intelligence Review, 54 0 (3): 0 1937–1967, Aug. 2020. ISSN 1573-7462. doi:10.1007/s10462-020-09896-5. URL http://dx.doi.org/10.1007/s10462-020-09896-5

doi:10.1007/s10462-020-09896-5 1937
[4]

Chen and C

T. Chen and C. Guestrin. Xgboost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD ’16. ACM, Aug. 2016. doi:10.1145/2939672.2939785. URL http://dx.doi.org/10.1145/2939672.2939785

arXiv 2016
[5]

Davenport and R

T. Davenport and R. Kalakota. The potential for artificial intelligence in healthcare. Future Healthcare Journal, 6 0 (2): 0 94--98, 2019. doi:10.7861/futurehosp.6-2-94

doi:10.7861/futurehosp.6-2-94 2019
[6]

Feurer and F

M. Feurer and F. Hutter. Hyperparameter Optimization, pages 3--33. Springer International Publishing, Cham, 2019. ISBN 978-3-030-05318-5. doi:10.1007/978-3-030-05318-5_1. URL https://doi.org/10.1007/978-3-030-05318-5_1

doi:10.1007/978-3-030-05318-5_1 2019
[7]

Feurer, J

M. Feurer, J. Springenberg, and F. Hutter. Initializing bayesian hyperparameter optimization via meta-learning. Proceedings of the AAAI Conference on Artificial Intelligence, 29 0 (1), Feb. 2015. doi:10.1609/aaai.v29i1.9354. URL https://ojs.aaai.org/index.php/AAAI/article/view/9354

doi:10.1609/aaai.v29i1.9354 2015
[8]

Friedman

M. Friedman. The use of ranks to avoid the assumption of normality implicit in the analysis of variance. Journal of the American Statistical Association, 32 0 (200): 0 675--701, 1937. doi:10.1080/01621459.1937.10503522. URL https://www.tandfonline.com/doi/abs/10.1080/01621459.1937.10503522

arXiv 1937

Show all 23 references

[9]

S. Holm. A simple sequentially rejective multiple test procedure. Scandinavian Journal of Statistics, 6: 0 65--70, 1979. URL https://api.semanticscholar.org/CorpusID:122415379

1979
[10]

Ismail Fawaz, G

H. Ismail Fawaz, G. Forestier, J. Weber, L. Idoumghar, and P.-A. Muller. Deep learning for time series classification: a review. Data Mining and Knowledge Discovery, 33 0 (4): 0 917--963, 2019

2019
[11]

H. S. Jomaa, L. Schmidt-Thieme, and J. Grabocka. Dataset2Vec : Learning Dataset Meta - Features . Data Mining and Knowledge Discovery, 35: 0 964--985, 2021. doi:https://doi.org/10.1007/s10618-021-00737-9

2021 doi
[12]

M. Kaya, M. Kaya, M. Kaya, H. Şakir Bilge, H. S. Bilge, and H. Şakir Bilge. Deep metric learning: A survey. Symmetry, 2019. doi:10.3390/sym11091066

2019 doi
[13]

J. Kim, S. Kim, and S. Choi. Learning to warm-start bayesian hyperparameter optimization, 2018. URL https://arxiv.org/abs/1710.06219

2018 arXiv
[14]

S. Lloyd. Least squares quantization in pcm. IEEE Transactions on Information Theory, 28 0 (2): 0 129--137, 1982. doi:10.1109/TIT.1982.1056489

1982
[15]

D. D. Mohan, B. Jawade, S. Setlur, and V. Govindaraj. Deep metric learning for computer vision: A brief overview, 2023. URL https://arxiv.org/abs/2312.10046

2023 arXiv
[16]

Probst, A.-L

P. Probst, A.-L. Boulesteix, and B. Bischl. Tunability: Importance of Hyperparameters of Machine Learning Algorithms . The Journal of Machine Learning Research, 20 0 (1): 0 1934--1965, 2019

1934
[17]

Płudowski, A

D. Płudowski, A. Zajko, A. Kozak, and K. Woźnica. Rethinking of encoder-based warm-start methods in hyperparameter optimization, 2024. URL https://arxiv.org/abs/2403.04720

2024 arXiv
[18]

Rakotoarison, L

H. Rakotoarison, L. Milijaona, A. Rasoanaivo, M. Sebag, and M. Schoenauer. Learning meta-features for automl. International Conference on Learning Representations, 2022. doi:null

2022
[19]

Turner, D

R. Turner, D. Eriksson, M. McCourt, J. Kiili, E. Laaksonen, Z. Xu, and I. Guyon. Bayesian optimization is superior to random search for machine learning hyperparameter tuning: Analysis of the black-box optimization challenge 2020, 2021. URL https://arxiv.org/abs/2104.10201

2020 arXiv
[20]

Vanschoren, J

J. Vanschoren, J. N. van Rijn, B. Bischl, and L. Torgo. OpenML : Networked Science in Machine Learning . SIGKDD Explorations, 15 0 (2): 0 49--60, 2013. doi:10.1145/2641190.2641198

2013
[21]

Wistuba, N

M. Wistuba, N. Schilling, and L. Schmidt-Thieme. Two-stage transfer surrogate model for automatic hyperparameter optimization. In P. Frasconi, N. Landwehr, G. Manco, and J. Vreeken, editors, Machine Learning and Knowledge Discovery in Databases, pages 199--214, Cham, 2016. Spr...

2016
[22]

Zaheer, S

M. Zaheer, S. Kottur, S. Ravanbakhsh, B. Poczos, R. R. Salakhutdinov, and A. J. Smola. Deep Sets . In Advances in Neural Information Processing Systems , volume 30, 2017

2017
[23]

J. H. Zar. Spearman rank correlation. Encyclopedia of Biostatistics, 7, 2005

2005

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [1]

A. Alanazi. Using machine learning for healthcare challenges and opportunities. Informatics in Medicine Unlocked, 30: 0 100924, 2022. doi:10.1016/j.imu.2022.100924

arXiv 2022

[2] [2]

T. Bai, Y. Li, Y. Shen, X. Zhang, W. Zhang, and B. Cui. Transfer learning for bayesian optimization: A survey, 2023. URL https://arxiv.org/abs/2302.05927

arXiv 2023

[3] [3]

Bentéjac, A

C. Bentéjac, A. Csörgő, and G. Martínez-Muñoz. A comparative analysis of gradient boosting algorithms. Artificial Intelligence Review, 54 0 (3): 0 1937–1967, Aug. 2020. ISSN 1573-7462. doi:10.1007/s10462-020-09896-5. URL http://dx.doi.org/10.1007/s10462-020-09896-5

doi:10.1007/s10462-020-09896-5 1937

[4] [4]

Chen and C

T. Chen and C. Guestrin. Xgboost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD ’16. ACM, Aug. 2016. doi:10.1145/2939672.2939785. URL http://dx.doi.org/10.1145/2939672.2939785

arXiv 2016

[5] [5]

Davenport and R

T. Davenport and R. Kalakota. The potential for artificial intelligence in healthcare. Future Healthcare Journal, 6 0 (2): 0 94--98, 2019. doi:10.7861/futurehosp.6-2-94

doi:10.7861/futurehosp.6-2-94 2019

[6] [6]

Feurer and F

M. Feurer and F. Hutter. Hyperparameter Optimization, pages 3--33. Springer International Publishing, Cham, 2019. ISBN 978-3-030-05318-5. doi:10.1007/978-3-030-05318-5_1. URL https://doi.org/10.1007/978-3-030-05318-5_1

doi:10.1007/978-3-030-05318-5_1 2019

[7] [7]

Feurer, J

M. Feurer, J. Springenberg, and F. Hutter. Initializing bayesian hyperparameter optimization via meta-learning. Proceedings of the AAAI Conference on Artificial Intelligence, 29 0 (1), Feb. 2015. doi:10.1609/aaai.v29i1.9354. URL https://ojs.aaai.org/index.php/AAAI/article/view/9354

doi:10.1609/aaai.v29i1.9354 2015

[8] [8]

Friedman

M. Friedman. The use of ranks to avoid the assumption of normality implicit in the analysis of variance. Journal of the American Statistical Association, 32 0 (200): 0 675--701, 1937. doi:10.1080/01621459.1937.10503522. URL https://www.tandfonline.com/doi/abs/10.1080/01621459.1937.10503522

arXiv 1937