Pith. sign in

REVIEW 4 major objections 4 minor 49 references

Post-hoc Interpretability Illumination for Scientific Interaction Discovery

T0 review · 4 major / 4 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Iterative Kings' Forests (iKF) is a post-hoc tree method that produces ranked, typed multi-order interaction candidates from a trained model without retraining.

desk verdict A real extension of iRF whose headline 'post-hoc for any model' claim is untested; the King-root idea and three-type taxonomy are worth debating, but the paper needs major revision, not rejection. read the letter →

arxiv 2412.16252 v1 pith:YCRGILV4 submitted 2024-12-20 cs.LG cs.AIstat.ML

Ling Zhang , Zhichao Hou , Tingxiang Ji , Yuanyuan Xu , Runze Li This is my paper · ORCID
classification cs.LGcs.AIstat.ML
keywords IterativeKings'Forestspost-hocinterpretabilityvariableinteractionsimportancerandomDrosophilaenhancersfeaturescreeningscientificdiscovery
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 proposes Iterative Kings' Forests (iKF), a post-hoc interpretability method that discovers multi-order interactions among variables. iKF repeatedly names one variable the "King," builds a forest in which every tree is rooted at that King, and uses a permutation-based importance measure, King's PVIM, to reweight variables and rank depth-specific tree paths as interaction candidates. On simulated functions with known pairwise and third-order ground truth, iKF recovers the active interactions more often than iterative random forests and ranks the active variables ahead of distance-correlation screening. On Drosophila embryo enhancer data, iKF rediscovers 14 of the 16 transcription-factor interactions previously verified in the literature and adds three further Zld-centred pairwise interactions and one third-order interaction that also have literature support. The authors claim this makes iKF usable on any trained model, without retraining, to generate ranked shortlists of interaction hypotheses and to label each as Accompanied, Synergistic, or Hierarchical.

What carries the argument

The load-bearing object is the King's Forest: a forest of trees with the chosen King variable fixed at every root. King's PVIM, defined as the change in prediction error when the King is permuted, is used both to select informative trees and to update variable weights, so the King's interaction partners accumulate weight and migrate toward the root. From the resulting forests, iKF extracts all depth-d root-to-leaf paths and ranks them by two metrics: summed King's PVIM and path reproduction count. The rule for assigning interaction type is that a substantial jump in King's PVIM from depth d to d+1 signals an order-(d+1) interaction, path-direction symmetry distinguishes Synergistic from Hierarchical interactions, and the presence of a marginal effect marks an Accompanied interaction. These ranked paths and type labels are the output a scientist would use to generate mechanistic hypotheses.

What would settle it

Generate data from a function with only pairwise interactions, $y = x_1 x_3 + x_5 x_7 + \text{noise}$, run iKF at maximum depth 3 over 100 replicates, and count how often a depth-3 path containing $x_1$, $x_3$, and an irrelevant variable appears in the top-ranked lists; if false triples are common, the depth-jump rule does not establish interaction order.

Watch

Extended reading notes

Core claim

The central claim is that forcing a chosen "King" variable to sit at the root of every tree converts a random forest into a targeted interaction detector: the variables that share a path with the King and are rewarded by King's PVIM are exactly the variables that interact with it, and the depth at which King's PVIM jumps reveals the interaction order. In simulations, iKF outperforms iRF in overall interaction recovery, for example 0.60 versus 0.09 overall recovery rate in case (a1) and 0.68 and 0.48 recovery in the third-order settings (b1) and (b2) where iRF recovers none. On the Drosophila data, iKF identifies Zld as a dominant transcription factor whose interactions with gap-gene and anteroposterior-patterning TFs are hierarchical, while the Zld-Twi interaction is synergistic, matching the documented biology that Zld licenses downstream genes. The paper therefore presents iKF as a general post-hoc scientific-discovery tool that returns ranked, typed candidate interactions rather than a single importance score.

Load-bearing premise

The method assumes that a sharp rise in a King's importance score when trees are allowed to grow deeper is a reliable sign of a true higher-order interaction, and that the model being scored has actually learned the interaction.

Editorial extensions

If this is right

  • A researcher can run iKF on a trained model and receive ranked shortlists of candidate variable interactions at several orders, with no retraining step.
  • The method attaches a type to each candidate interaction, so a biologist can immediately separate synergistic pairs from nested hierarchical regulation.
  • In simulations, iKF finds third-order interactions that the iRF baseline never recovers, suggesting it extends interaction discovery beyond pairwise effects.
  • On Drosophila enhancer data, iKF recovers known Zld-centred regulatory interactions and identifies Zld as the dominant factor, matching the experimentally documented hierarchy.
  • Because Eq. (2) is defined on predictions of any given model, iKF could provide the same interaction discovery for neural networks and other black boxes if the post-hoc claim holds.

Reading between the lines

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

  • The paper's own caveat is that iKF cannot identify all interactions, and its (b3) simulation with p=500, where the all-active screening rate at model size d2 is only 0.08, is a concrete reminder that the output should be read as ranked hypotheses rather than a complete census.
  • A sharper version of iKF would replace the qualitative "sharp increase" rule for interaction order with a significance threshold or bootstrap interval over replications, which the 100-replication simulation protocol makes straightforward to build.
  • The input to Algorithm 1 never names the predictor model, and the experiments appear to use the King's forests themselves; applying iKF to a fixed black-box model's predictions is the direct test of the strongest "any model" reading.
  • Path-direction asymmetry is a cheap, interpretable signature of hierarchical interactions that could be benchmarked directly on synthetic functions whose true generative order is known.
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.

Referee Report

4 major / 4 minor

Summary. The paper proposes Iterative Kings' Forests (iKF), a method that, given a designated 'King' variable, builds forests with that variable fixed at the root, uses King's permutation variable importance to reweight variables, and outputs ranked depth-d path shortlists together with three interaction types: Accompanied, Synergistic, and Hierarchical. The authors claim that iKF is a post-hoc interpretability tool applicable to any model without retraining, and they evaluate it on simulated functions against DC-SIS and iRF, reporting lower minimum recovery sizes and higher interaction recovery rates. They also apply iKF to Drosophila enhancer data and claim rediscovery of biologically verified transcription-factor interactions involving Zld.

Significance. If the post-hoc claim were supported, iKF would be a convenient tool for extracting multi-order interaction candidates from arbitrary trained models, with clear value for genomics and other scientific domains. The paper has genuine strengths: it compares against two relevant baselines across numerous simulation settings, reports quantile-based recovery sizes and recovery rates, and grounds the Drosophila findings in specific biological literature. However, the central claim is not supported by the algorithm as written or by the experiments, and the simulation protocol is internally inconsistent, so the significance of the contribution is not currently established.

major comments (4)
  1. [§3.2, Eq. (2), Algorithm 1] Equation (2) defines PVIM using predictions ŷ_i of 'any given model', but Algorithm 1 never accepts or names such a model, and no experiment states which model produced the predictions used in Eq. (2). The only coherent reading is that the King's forests themselves predict the response, which makes iKF a standalone random-forest-based screening method rather than a post-hoc explainer for an arbitrary trained model. The abstract's and Section 1's central claim that iKF 'can be applied to any model ... without retraining' is therefore untested, and the per-tree King's PVIM used in Eq. (3) is not even defined when the predictor is a separate global model. The paper needs either an explicit model input with a model-agnostic PVIM definition and experiments on a separately trained model, or a reframing of the contribution; as written, the advertised post-hoc property is unsupported.
  2. [§4.1 vs Appendix A] The simulation ground-truth functions are not stated consistently. Section 4.1 defines (a2) as y = 4 x1 sin(x3) − 4 x5 cos(x7) and (b2) as y = 2 x1 sign(1 + x3) sin(x5), while Appendix A defines (a2) as y = 2s x1 sin(x3) + 2s x5 cos(x7 + π/2) and (b2) as y = s x1 log(5|1+x3|) sin(x5). The appendix also has five settings per part while Section 4.1 has three, and the scaling parameter s is never defined. Consequently the ORR and IRR numbers in Table 1 cannot be tied to a uniquely specified simulation, and the experiments are not reproducible as written.
  3. [§4.1, Algorithm 2] The selection of the first King x(1) is an input to Algorithm 2, but the experimental protocol never reports how it was chosen in the simulations or in the Drosophila study. Section 3.2 allows a random choice when no prior knowledge is available, but the results in Figure 2 and Tables 1, 9–15 contain no seed or initialization details. Since every subsequent King and weight update depends on x(1), this is a load-bearing underspecification of the method.
  4. [§3.3, Tables 2, 4, 7, 8] The criterion for interaction order and type is a 'sharp increase' in King's PVIM across depths, but no threshold, null distribution, or replication-based uncertainty is given. In Table 2, x7's PVIM rises from −0.21 to 1.79 and is treated as evidence of an interaction, while x10's rise from −0.19 to 0.33 is dismissed; in Table 4, x1's change from 1.11 to 1.33 is described as 'keeping increasing'. In Section 5.3, conclusions about Synergistic versus Hierarchical interactions separate PVIM values around 0.02 from values near 1e−16 with no error bars or formal comparison. The interaction-type labels therefore rest on an unstated decision rule rather than a defined statistical criterion.
minor comments (4)
  1. [§4.2] The text says 'a smaller S indicates greater recovery power'; the definition just introduced calls this quantity MRS, and S is later used for the survived variable set, so the notation should be corrected.
  2. [§4.3, Case 2] The text states that PVIMs of 'variables 1 and 5' increase sharply and then says 'This indicates that 3 and 5 are involved'; the latter should refer to variables 1 and 5.
  3. [§4.1] The hyperparameter Nc is written as 'Nc =⌊n/ 2 log(n)⌋', which is ambiguous between n/(2 log n) and (n/2) log n; the intended formula should be disambiguated.
  4. [General] No code or data release is mentioned; given the many simulation tables and the need to reconstruct the exact protocol, a reproducibility statement or code link would be needed in any revision.

Circularity Check

1 steps flagged · score 4.0 of 10

Post-hoc 'any model' claim is self-referential because Algorithm 1 never accepts a model, so the King's PVIMs and path lists explain iKF's own forests rather than a separately given predictor; the interaction-discovery results, however, are independently benchmarked.

  1. self definitional [Eq. (2), Section 3.1; Algorithm 1, Section 3.2; Section 4.2 and Section 5.3 experiments]
    "PVIM = (sum_i (yi - yhat*_i)^2 - sum_i (yi - yhat_i)^2) / |B| ... yhat_i and yhat*_i are the predictions of any given model for data i before and after permuting the selected variable. ... Algorithm 1: Input: Forest size N, maximum depth D, King."

    The paper's central claim is that iKF is a post-hoc tool applicable to any model without retraining. But Algorithm 1's input does not include a model; it only takes forest size, maximum depth, and a King. The PVIM in Eq. (2), which drives all weight updates and path ranking, is defined using predictions of 'any given model', yet no external model is specified anywhere in Algorithms 1-2 or in the experiments. In the simulations and the Drosophila case study, the predictions are evidently produced by the King's Forests that iKF itself constructs. Thus the object being interpreted and the interpreter are the same forest: the interaction evidence (King's PVIMs, path lists, interaction types) is a property of iKF's own fit, not of an independently trained black-box model.

full rationale

The only substantive circularity I can exhibit with the paper's own equations is the self-referential use of iKF's own forests as the 'any given model' in PVIM. This undermines the post-hoc, model-agnostic framing in the abstract and Section 1, because Algorithm 1 never accepts a model and the experiments never identify an external predictor. The interaction-discovery machinery, however, is not circular in the narrower statistical sense: iKF is a data-driven iterative reweighting procedure, and its outputs are tested against simulated ground-truth functions and against externally documented Drosophila transcription-factor interactions (Tables 16-17). Those comparisons give independent evidence that the method recovers real interactions, even if the 'post-hoc any model' claim is unsupported. The apparent inconsistency between the simulation formulas in Section 4.1 and Appendix A (e.g., (a2) and (b2) differ in functional form) is a reproducibility concern, not a circularity. No load-bearing self-citation or imported uniqueness theorem appears. The interaction-type labels rely on heuristic thresholds for PVIM jumps, but that is a statistical robustness concern rather than a reduction of the result to its inputs. Overall, the central interaction-discovery claim has independent empirical grounding, so the circularity score is moderate, reflecting the self-referential post-hoc framing rather than a fully forced derivation.

Assumptions & free parameters 6 free parameters · 5 assumptions · 3 invented entities

The central method rests on several unproved heuristics (paths equal interactions, PVIM jumps equal interaction order, direction equals dominance) and on an unspecified predictive model. The only variables chosen by hand are hyperparameters; the paper introduces conceptual entities (King, Core Team, interaction types) that lack independent evidence beyond the paper's own simulations and selected literature checks.

free parameters (6)
  • Survival fraction alpha = 0.5 (simulations), 0.2 (Drosophila)
    Chosen by hand in Sections 4.1 and 5.1 to control how many variables survive each iKF iteration; affects which variables can become Kings.
  • Candidate team size Nc = floor(p/2) (Drosophila), floor(n/(2 log n)) (simulations)
    Controls the number of variables eligible for core-team construction; set per experiment without sensitivity analysis.
  • Number of weight iterations Niter = 7 (simulations), 6 (Drosophila)
    Stopping of the inner weight update; no convergence criterion or sensitivity analysis is given.
  • Maximum depth D = 4 or 5 (simulations), 5 (Drosophila)
    Determines the highest interaction order searchable; set by the known ground truth in simulations, ad hoc in biology.
  • Shortlist size Ntop = 20 (simulations), 30 (Drosophila)
    Number of candidate paths reported per depth; an arbitrary threshold that defines which interactions are called discovered.
  • Stopping threshold K = Not specified
    Algorithm 2 stops when the survived variable set falls below K, but K is never reported in either experiment.
assumptions (5)
  • domain assumption A depth-d tree path represents a potential d-order interaction among the variables on that path.
    Section 3.1 and Algorithm 1 use depth-d paths as interaction candidates without proving that co-occurrence in a path implies interaction.
  • domain assumption A sharp increase in King's PVIM from depth d to d+1 indicates the King participates in an order-(d+1) interaction.
    Section 3.3 states this criterion; no formal threshold or proof is given.
  • domain assumption Both directions of a path appearing in shortlists means equal-status variables, while one direction means hierarchical nesting.
    Section 3.3 defines path directions; this ordering interpretation is assumed, not derived.
  • domain assumption Variables and error are independent standard normal in simulations.
    Section 4.1 assumes standard normal xi and error e; independence is implied but not stated, and real data may violate it.
  • domain assumption The model used for permutation importance is the King's forest itself, or an unspecified black box with accurate predictions.
    Eq. (2) requires a predictive model, but Algorithm 1 never specifies which model supplies predictions; the experiments appear to use the iKF-built forests.
invented entities (3)
  • King variable
    purpose: A user-selected or weight-selected variable placed at every tree root to anchor interaction search.
    It is a procedural construct with no independent falsifiable handle; its utility is measured only inside the paper's simulations.
  • King's Core Team
    purpose: The set of variables that share at least one interaction with the King.
    Defined in Sections 3.1-3.2; membership is inferred from iKF's own weights and path lists, not from an external criterion.
  • Three interaction types (Accompanied, Synergistic, Hierarchical)
    purpose: Taxonomy to label discovered interactions for scientists.
    The categories are defined by the authors in Section 3.3 and assigned heuristically from PVIM patterns; specific biological assignments in Section 5.3 have literature support, but the taxonomy itself is not independently corroborated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Post-hoc Interpretability Illumination for Scientific Interaction Discovery." pith.science (2026). https://pith.science/paper/YCRGILV4

@misc{pith2026241216252,
  author       = {Pith},
  title        = {Pith review of: Post-hoc Interpretability Illumination for Scientific Interaction Discovery},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YCRGILV4}},
  note         = {Machine review of arXiv:2412.16252}
}
read the original abstract

Model interpretability and explainability have garnered substantial attention in recent years, particularly in decision-making applications. However, existing interpretability tools often fall short in delivering satisfactory performance due to limited capabilities or efficiency issues. To address these challenges, we propose a novel post-hoc method: Iterative Kings' Forests (iKF), designed to uncover complex multi-order interactions among variables. iKF iteratively selects the next most important variable, the "King", and constructs King's Forests by placing it at the root node of each tree to identify variables that interact with the "King". It then generates ranked short lists of important variables and interactions of varying orders. Additionally, iKF provides inference metrics to analyze the patterns of the selected interactions and classify them into one of three interaction types: Accompanied Interaction, Synergistic Interaction, and Hierarchical Interaction. Extensive experiments demonstrate the strong interpretive power of our proposed iKF, highlighting its great potential for explainable modeling and scientific discovery across diverse scientific fields.

Figures

Figures reproduced from arXiv: 2412.16252 by the authors.

Figure 1
Figure 1. iKF Overview: Iteratively select new Kings until [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Minimum recovery size (MRS) across various quanti [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 41 canonical work pages

  1. [1]

    Cabrera, and Y.-S

    Amaratunga, D., J. Cabrera, and Y.-S. Lee (2008). Enriched random forests. Bioinformatics\/ 24\/ (18), 2010--2014

    work page 2008

  2. [2]

    Ceolini, C

    Ancona, M., E. Ceolini, C. \"O ztireli, and M. Gross (2017). Towards better understanding of gradient-based attribution methods for deep neural networks. arXiv preprint arXiv:1711.06104\/

    arXiv 2017

  3. [3]

    Archer, K. J. and R. V. Kimes (2008). Empirical characterization of random forest variable importance measures. Computational Statistics & Data Analysis\/ 52\/ (4), 2249--2260

    work page 2008

  4. [4]

    Kumbier, J

    Basu, S., K. Kumbier, J. B. Brown, and B. Yu (2018). Iterative random forests to discover predictive and stable high-order interactions. Proceedings of the National Academy of Sciences\/ , 201711236

    work page 2018

  5. [5]

    Breiman, L. (2001). Random forests. Machine learning\/ 45\/ (1), 5--32

    work page 2001

  6. [6]

    Capovilla, M., E. D. Eldon, and V. Pirrotta (1992). The giant gene of drosophila encodes a b-zip dna-binding protein that regulates the expression of other segmentation gap genes. Development\/ 114\/ (1), 99--112

    work page 1992

  7. [7]

    D \' az-Uriarte, R. and S. A. De Andres (2006). Gene selection and classification of microarray data using random forest. BMC bioinformatics\/ 7\/ (1), 3

    work page 2006

  8. [8]

    Eldon, E. D. and V. Pirrotta (1991). Interactions of the drosophila gap gene giant with maternal and zygotic pattern-forming genes. Development\/ 111\/ (2), 367--378

    work page 1991

Show all 49 references
  1. [9]

    Elmarakeby, H. A., J. Hwang, R. Arafeh, J. Crowdis, S. Gang, D. Liu, S. H. AlDubayan, K. Salari, S. Kregel, C. Richter, et al. (2021). Biologically informed deep neural network for prostate cancer discovery. Nature\/ 598\/ (7880), 348--352

    2021

  2. [10]

    Fan, J. and R. Li (2001). Variable selection via nonconcave penalized likelihood and its oracle properties. Journal of the American statistical Association\/ 96\/ (456), 1348--1360

    2001

  3. [11]

    Fan, J. and J. Lv (2008). Sure independence screening for ultrahigh dimensional feature space. Journal of the Royal Statistical Society: Series B (Statistical Methodology)\/ 70\/ (5), 849--911

    2008

  4. [12]

    M., X.-Y

    Harrison, M. M., X.-Y. Li, T. Kaplan, M. R. Botchan, and M. B. Eisen (2011). Zelda binding in the early drosophila melanogaster embryo marks regions subsequently activated at the maternal-to-zygotic transition. PLoS genetics\/ 7\/ (10), e1002266

    2011

  5. [13]

    o der, E. Seifert, and H. J \

    Hoch, M., C. Schr \"o der, E. Seifert, and H. J \"a ckle (1990). cis-acting control elements for kr \"u ppel expression in the drosophila embryo. The EMBO journal\/ 9\/ (8), 2587--2595

    1990

  6. [14]

    a ckle (1991). Gene expression mediated by cis-acting sequences of the kr \

    Hoch, M., E. Seifert, and H. J \"a ckle (1991). Gene expression mediated by cis-acting sequences of the kr \"u ppel gene in response to the drosophila morphogens bicoid and hunchback. The EMBO journal\/ 10\/ (8), 2267--2278

    1991

  7. [15]

    Ghashami, M

    Hou, Z., M. Ghashami, M. Kuznetsov, and M. Torkamani (2024). Hlogformer: A hierarchical transformer for representing log data. arXiv preprint arXiv:2408.16803\/

    2024 arXiv

  8. [16]

    Hou, Z., J. Leng, J. Yu, Z. Xia, and L.-Y. Wu (2023). Pathexpsurv: pathway expansion for explainable survival analysis and disease gene discovery. BMC bioinformatics\/ 24\/ (1), 434

    2023

  9. [17]

    Hou, Z., M. Lin, M. Torkamani, S. Wang, and X. Liu (2024). Adversarial robustness in graph neural networks: Recent advances and new frontier. In 2024 IEEE 11th International Conference on Data Science and Advanced Analytics (DSAA) , pp.\ 1--2. IEEE

    2024

  10. [18]

    Kraut, R. and M. Levine (1991). Spatial regulation of the gap gene giant during drosophila development. Development\/ 111\/ (2), 601--609

    1991

  11. [19]

    Levine, M. (2010). Transcriptional enhancers in animal development and evolution. Current Biology\/ 20\/ (17), R754--R763

    2010

  12. [20]

    Zhong, and L

    Li, R., W. Zhong, and L. Zhu (2012). Feature screening via distance correlation learning. Journal of the American Statistical Association\/ 107\/ (499), 1129--1139

    2012

  13. [21]

    MacArthur, R

    Li, X.-y., S. MacArthur, R. Bourgon, D. Nix, D. A. Pollard, V. N. Iyer, A. Hechmer, L. Simirenko, M. Stapleton, C. L. L. Hendriks, et al. (2008). Transcription factors bind thousands of active and inactive regions in the drosophila blastoderm. PLoS biology\/ 6\/ (2), e27

    2008

  14. [22]

    Nien, H.-Y

    Liang, H.-L., C.-Y. Nien, H.-Y. Liu, M. M. Metzstein, N. Kirov, and C. Rushlow (2008). The zinc-finger protein zelda is a key activator of the early zygotic genome in drosophila. Nature\/ 456\/ (7220), 400

    2008

  15. [23]

    Lin, T., Y. Wang, X. Liu, and X. Qiu (2022). A survey of transformers. AI open\/ 3 , 111--132

    2022

  16. [24]

    Lundberg, S. (2017). A unified approach to interpreting model predictions. arXiv preprint arXiv:1705.07874\/

    2017 arXiv

  17. [25]

    Ma, M. and R. Fergus (2013). Visualizing and understanding convolutional networks. Computer Vision--ECCV 2014\/

    2013

  18. [26]

    Zinzen, P

    Markstein, M., R. Zinzen, P. Markstein, K.-P. Yee, A. Erives, A. Stathopoulos, and M. Levine (2004). A regulatory code for neurogenic gene expression in the drosophila embryo. Development\/ 131\/ (10), 2387--2394

    2004

  19. [27]

    Mor \'a n, \'E . and G. Jim \'e nez (2006). The tailless nuclear receptor acts as a dedicated repressor in the early drosophila embryo. Molecular and Cellular Biology\/ 26\/ (9), 3446--3454

    2006

  20. [28]

    Nguyen, H. T. and X. Xu (1998). Drosophila mef2expression during mesoderm development is controlled by a complex array ofcis-acting regulatory modules. Developmental biology\/ 204\/ (2), 550--566

    1998

  21. [29]

    Liang, S

    Nien, C.-Y., H.-L. Liang, S. Butcher, Y. Sun, S. Fu, T. Gocha, N. Kirov, J. R. Manak, and C. Rushlow (2011). Temporal coordination of gene networks by zelda in the early drosophila embryo. PLoS genetics\/ 7\/ (10), e1002339

    2011

  22. [30]

    Ribeiro, M. T., S. Singh, and C. Guestrin (2016). Explaining the predictions of any classifier. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining , pp.\ 1135--1144

    2016

  23. [31]

    Rivera-Pomar, R. and H. Jackle (1996). From gradients to stripes in drosophila embryogenesis: filling in the gaps. Trends in Genetics\/ 12\/ (11), 478--483

    1996

  24. [32]

    Schulz, C. and D. Tautz (1994). Autonomous concentration-dependent activation and repression of kruppel by hunchback in the drosophila embryo. Development\/ 120\/ (10), 3043--3049

    1994

  25. [33]

    Shah, R. D. (2016). Modelling interactions in high-dimensional data with backtracking. Journal of Machine Learning Research\/ 17\/ (207), 1--31

    2016

  26. [34]

    Shah, R. D. and N. Meinshausen (2014). Random intersection trees. The Journal of Machine Learning Research\/ 15\/ (1), 629--654

    2014

  27. [35]

    Greenside, and A

    Shrikumar, A., P. Greenside, and A. Kundaje (2017). Learning important features through propagating activation differences. In International conference on machine learning , pp.\ 3145--3153. PMlR

    2017

  28. [36]

    Friedman, T

    Simon, N., J. Friedman, T. Hastie, and R. Tibshirani (2013). A sparse-group lasso. Journal of computational and graphical statistics\/ 22\/ (2), 231--245

    2013

  29. [37]

    Van Drenth, A

    Stathopoulos, A., M. Van Drenth, A. Erives, M. Markstein, and M. Levine (2002). Whole-genome analysis of dorsal-ventral patterning in the drosophila embryo. Cell\/ 111\/ (5), 687--701

    2002

  30. [38]

    Johnston, and P

    Struhl, G., P. Johnston, and P. A. Lawrence (1992). Control of drosophila body pattern by the hunchback morphogen gradient. Cell\/ 69\/ (2), 237--249

    1992

  31. [39]

    Chen, T.-J

    Sze, V., Y.-H. Chen, T.-J. Yang, and J. S. Emer (2017). Efficient processing of deep neural networks: A tutorial and survey. Proceedings of the IEEE\/ 105\/ (12), 2295--2329

    2017

  32. [40]

    Tibshirani, R. (1996). Regression shrinkage and selection via the lasso. Journal of the Royal Statistical Society Series B: Statistical Methodology\/ 58\/ (1), 267--288

    1996

  33. [41]

    Ma, H.-C

    Xu, H., Y. Ma, H.-C. Liu, D. Deb, H. Liu, J.-L. Tang, and A. K. Jain (2020). Adversarial attacks and defenses in images, graphs and text: A review. International journal of automation and computing\/ 17 , 151--178

    2020

  34. [42]

    Zhang, R

    Yang, G., L. Zhang, R. Li, and Y. Huang (2019). Feature screening in ultrahigh-dimensional varying-coefficient cox model. Journal of Multivariate Analysis\/ 171\/ (C), 284--297

    2019

  35. [43]

    Yuan, M. and Y. Lin (2006). Model selection and estimation in regression with grouped variables. Journal of the Royal Statistical Society Series B: Statistical Methodology\/ 68\/ (1), 49--67

    2006

  36. [44]

    Zeitlinger, J., R. P. Zinzen, A. Stark, M. Kellis, H. Zhang, R. A. Young, and M. Levine (2007). Whole-genome chip--chip analysis of dorsal, twist, and snail suggests integration of diverse patterning processes in the drosophila embryo. Genes & development\/ 21\/ (4), 385--390

    2007

  37. [45]

    Zhao, L., Q. Dong, C. Luo, Y. Wu, D. Bu, X. Qi, Y. Luo, and Y. Zhao (2021). Deepomix: a scalable and interpretable multi-omics deep learning framework and application in cancer survival analysis. Computational and structural biotechnology journal\/ 19 , 2719--2725

    2021

  38. [46]

    Zhou, J. and O. G. Troyanskaya (2015). Predicting effects of noncoding variants with deep learning--based sequence model. Nature methods\/ 12\/ (10), 931--934

    2015

  39. [47]

    Zintgraf, L. M., T. S. Cohen, T. Adel, and M. Welling (2017). Visualizing deep neural network decisions: Prediction difference analysis. arXiv preprint arXiv:1702.04595\/

    2017 arXiv

  40. [48]

    Zou, H. and T. Hastie (2005). Regularization and variable selection via the elastic net. Journal of the Royal Statistical Society Series B: Statistical Methodology\/ 67\/ (2), 301--320

    2005

  41. [49]

    Zou, H. and R. Li (2008). One-step sparse estimates in nonconcave penalized likelihood models. Annals of statistics\/ 36\/ (4), 1509--1533

    2008

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.