Pith. sign in

REVIEW 4 major objections 6 minor 24 references

LDACP: Long-Delayed Ad Conversions Prediction Model for Bidding Strategy

T0 review · 4 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read For long-delayed ad conversions, the paper proposes LDACP, a two-module model that predicts campaign-level conversion counts so automated bidding can use the true expected CPA instead of the stale real-time count.

desk verdict The paper's offline metrics break for the zero-conversion campaigns it is meant to fix; fix that and the industrial story holds. read the letter →

arxiv 2411.16095 v1 pith:STWTRK6J submitted 2024-11-25 cs.LG

Peng Cui (1) , Yiming Yang (2) , Fusheng Jin (1) , Siyuan Tang (2) , Yunli Wang (2) , Fukang Yang (2) , Yalong Jia (2) , Qingpeng Cai (2)
show 4 more authors
Fei Pan (2) Changcheng Li (2) Peng Jiang (2) ((1) Beijing Institute of Technology (2) Kuaishou Technology)
This is my paper · ORCID
classification cs.LG
keywords adconversionspredictionlabelsmoothinglong-taildistributionautomatedbiddingstrategyonlineadvertisingcostcompliancerateproxylabelsmixtureofexperts
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

For online ads whose conversions arrive hours or days after the impression, the bidding system's real-time conversion count lags behind, so it overestimates the current cost per action and bids too timidly. This paper proposes predicting the final conversion count for each ad campaign at bidding-adjustment time, then feeding that predicted number into the bidding loop. The proposed model, LDACP, combines a bucket-classification module with smoothed labels and a value-regression module trained on proxy labels derived from the ranking model's prediction bias. In offline tests it beats the best baseline by 0.0686 in MAPE and 8.52 percentage points in cost-compliance rate; an online A/B test on the platform reports a 2.29% higher compliance rate, 11.06% higher platform revenue, and 8.68% more conversions.

What carries the argument

Two sub-modules carried by two distributional observations. BCMS (Bucket Classification Module with label Smoothing) builds on a tree-based bucket classifier but assigns each edge a soft target that varies continuously with distance to the bucket boundary, trained by cross-entropy for hard labels and MSE for soft labels. VRMP (Value Regression Module with Proxy labels) learns $\mathrm{PCOC}=z/y$, the ratio of aggregated pCTCVR to actual conversions, then recovers the predicted count as $\hat{y}_g = z/\hat{\mathrm{PCOC}}$. The MoE gate $\lambda$ is trained with a MAPE loss and shifts weight from the bucket module to the proxy-regression module as the label grows.

What would settle it

Compute the PCOC distribution on a different advertising platform or a different time window from the one used here. If many campaigns show PCOC values far from 1 or a heavy right tail, then VRMP's proxy-label regression loses its advantage; an offline comparison where LDACP fails to beat CREAD-P or TPM-N on such data would falsify the general claim.

Watch

Extended reading notes

Core claim

The paper's central claim is that campaign-level conversion counting for long-delayed ads is best done by splitting the task: a bucket classifier whose one-hot labels near bucket boundaries are replaced by smoothly varying soft labels, and a regression head that predicts the ranking model's bias (aggregated predicted conversion rate divided by actual conversions) rather than the raw long-tailed count. Because that bias ratio concentrates in a narrow range with no long tail on the authors' data, ordinary regression works for it, and converting back with the aggregated pCTCVR yields a tail-accurate count. A mixture-of-experts gate then blends the two predictions, weighting the bucket output for small counts and the proxy-regression output for large counts. The result is a single predicted conversion number that the automated bidding system can use in place of the incomplete real-time count.

Load-bearing premise

The whole proxy-regression side rests on an empirical claim: the ranking model's bias ratio PCOC stays in a narrow range with no long tail (Figure 2b on the offline dataset). If that ratio is heavy-tailed or has large outliers on another platform or period, converting back through it will be unstable and the long-tail problem returns.

Editorial extensions

If this is right

  • Bidding strategies can react to predicted final conversion counts rather than the small real-time count, avoiding the conservative bids that starve young campaigns of traffic.
  • Campaigns with conversion delays of hours to days (e.g., app installs, live-stream actions) get closer to their target CPA, raising the platform's compliance rate.
  • The platform's revenue and total conversions rise when inactive campaigns resume winning auctions.
  • The label-smoothing recipe is not tied to the specific tree classifier; the paper states it applies to any bucket-classification method with one-hot labels.
  • Tail prediction improves because the proxy-regression side handles large counts that the wide final bucket would otherwise over- or under-estimate.

Reading between the lines

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

  • If the narrow-range PCOC property turns out to be universal across platforms, the proxy-label trick could be reused for any delayed-feedback count where a biased online estimator exists.
  • The soft-label smoothing scheme could be tested on other integer-valued long-tail regression tasks, such as watch-time or purchase-count prediction, where boundary discontinuities hurt hard-label classifiers.
  • A natural stress test is to hold out product categories with unusually high or low PCOC spread; if VRMP degrades there, the MoE gate could be extended to use PCOC spread as an input feature.
  • The paper's five-day online test leaves open whether the gains persist over a longer horizon; measuring compliance rate and revenue drift over several weeks would be a direct follow-up.
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 / 6 minor

Summary. The paper addresses the problem of predicting the number of long-delayed ad conversions at the campaign level for use in automated bidding. The proposed LDACP model combines two modules: BCMS, which converts regression into a tree-structured bucket classification with non-normalized soft labels and label smoothing, and VRMP, which regresses the ratio PCOC = z/y (aggregated pCTCVR over actual conversions) and converts the prediction back to a conversion count. The two predictions are fused by a learned Mixture-of-Experts gate. The authors report offline experiments on a 2.7M-sample Kuai-AD dataset showing consistent improvements over six baselines (MAPE 0.2228 vs. 0.2914 for the best baseline; CR 62.32% vs. 53.80%), an ablation study attributing large gains to both modules, and an online A/B test on Kuaishou reporting a 2.29% compliance-rate improvement, 11.06% revenue increase, and 8.68% conversion increase.

Significance. If the reported results hold, the paper makes a useful industrial contribution: it formulates a practically important problem, provides a clean decomposition of the long-tail and label-discontinuity difficulties, and validates the method on a large real dataset and in an online deployment. The offline evaluation is substantial and the ablations are informative. I agree with the reader that the VRMP proxy is not circular: PCOC is a reparameterization of the target y, not a self-referential fitted constant. That said, the central offline claims rest on metrics that are not well-defined for y=0 samples, and the online test is reported without any statistical uncertainty. Both issues need to be resolved before the headline numbers can be taken at face value.

major comments (4)
  1. [Section 5.1.2, Eq. (19); Section 3, Table 2; Section 4.3, Eqs. (12) and (17)] The label space includes y=0 (Table 2 reports minimum conversions = 0), but the evaluation metrics CR in Eq. (19), MAPE in Section 5.1.2, and the MoE loss in Eq. (17) all divide by y. The paper does not state whether y=0 samples are excluded, assigned a special value, or handled by clipping. If they are dropped, the reported MAPE and CR describe only campaigns with at least one conversion, which is not the population most relevant to the bidding system: inactive campaigns with zero or very few tracked conversions are exactly where the predicted count is needed. If they are kept, the numerical results depend on an undisclosed convention and are not reproducible. This is load-bearing because Tables 3 and 4, including the headline MAPE 0.2228 and CR 62.32%, are computed from these metrics.
  2. [Section 4.3, Eq. (12); Figure 2b; Table 4] The core premise of VRMP is that PCOC has a narrow value range and no long tail (Section 4.3, Figure 2b). This property is in part manufactured by the definition in Eq. (12), which sets PCOC=1 for all y=0 samples even though z/y is undefined there. Without reporting the fraction of y=0 samples and the PCOC distribution conditional on y>0, Figure 2b cannot be read as evidence of a transferable distributional property. Since VRMP is credited with the largest ablation gain (CR drops by 9.42% when removed, Table 4), the authors should provide a more careful distributional analysis, including the zero-inflation rate, and state explicitly how the proxy is defined for zero-label samples in both training and evaluation.
  3. [Section 5.3, Table 5] The online A/B test is reported only as point estimates: the experimental group improves CR by 2.29% overall, with daily improvements ranging from 0.59 to 4.66 percentage points. No confidence intervals, p-values, number of campaigns, or traffic volumes are given, so the reader cannot assess whether the observed differences are statistically significant. The same applies to the revenue increase of 11.06% and the conversion increase of 8.68%. Given the commercial importance of the claim, the authors should report at least campaign-level variance and a significance test, or clearly label the result as a directional observation.
  4. [Section 5.1.1; Appendix A] The label window of three days is chosen without sensitivity analysis, and the conversion-delay statistics in Appendix A show that the p90 delay exceeds 2,770 minutes for several campaign types, so a three-day window still censors a non-negligible tail. Since all offline metrics and the downstream bidding signal depend on this label definition, the authors should either justify the window with data or provide results for alternative windows (e.g., one, two, four, and seven days).
minor comments (6)
  1. [Section 5.1.2, Eq. (19)] There is a typographical issue in Eq. (19): the summation is written as "Y∑︁ y", which should presumably be a standard sum over all labels y in the test set.
  2. [Section 4.3, Eq. (15)] The conversion from predicted PCOC to predicted conversion count divides by \hat{PCOC}, but the paper does not describe how zero or negative predictions of PCOC are handled; MAE regression can in principle produce such values, and the resulting division would be undefined or nonsensical.
  3. [Section 4.2.2, Eq. (9) and Table 1] The symbol e_i is used for the expectation value of a leaf node, but e_{i->j} is already used for an edge in Table 1; using two different meanings for e in the same paper is confusing and should be resolved by renaming one of them.
  4. [Section 5.1.1 and Section 4.3] The paper does not report the fraction of y=0 samples in Kuai-AD; this fraction is important both for interpreting the PCOC distribution in Figure 2b and for assessing how much of the reported CR/MAPE is determined by the convention for zero labels.
  5. [Section 5.3 and Section 4.4] The online deployment clips the predicted conversion count to be not less than the real-time tracked count, but the paper does not state whether this same clipping is applied to the offline predictions and to the MoE loss in Eq. (17); the offline and online settings should be aligned.
  6. [Section 5.1.3 and Eq. (18)] The hyperparameters alpha and beta in Eq. (18), as well as the smoothing temperature 10 in Eq. (21), are fixed empirically to 1 and 10, respectively, but no sensitivity study is reported; a short sensitivity analysis would increase confidence in the robustness of the method.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the VRMP proxy label is a learned ratio reparameterization, and the only self-citation (TPM) is code-backed and used as an empirical baseline.

full rationale

The derivation chain is self-contained. VRMP labels campaigns by PCOC = z/y (Eq. 12) and reconstructs predictions as y_g = z/PCOC_hat (Eq. 15); this is a bijective reparameterization of the target y given z, not a tautology, because PCOC_hat is learned from features x and z under the MAE loss (Eq. 14) and must generalize to held-out labels. BCMS and the MoE similarly minimize label-dependent losses (Eqs. 5, 17, 18) rather than returning fitted constants, so the reported offline gains (Table 3) and ablation gains (Table 4) are empirical, not by-construction. The only overlapping-author citation is TPM [14], used as the BCMS backbone and as a baseline; it is cited for its implementation ('we set the tree height to 7... using the implementation provided in the original paper') and is directly compared against in Table 3, so it is independent, code-reproduced support and does not import an unverified uniqueness claim. A caveat that is not circularity: with min conversions = 0 (Table 2), Eq. 12 sets PCOC=1 and Eqs. 17 and 19 divide by y, so CR/MAPE for zero-conversion samples depend on an undisclosed convention; this is a correctness/reproducibility risk, not self-reference.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The model introduces no new physical or conceptual entities; it combines existing ML techniques. The listed free parameters and axioms are the main assumptions the central claim depends on, mostly arising from ad hoc smoothing choices and the dataset-specific PCOC distribution.

free parameters (5)
  • alpha and beta in total loss = 1
    Loss balancing weights L_TOTAL = L_BCMS + alpha*L_VRMP + beta*L_MoE are set to 1 without sensitivity analysis (Section 4.4, Section 5.1.3). Performance could depend on these values.
  • Smoothing temperature in h(x)=e^{-10x} = 10
    The soft-label probability mapping uses a hand-chosen constant 10 in the exponent (Section 4.3, Eq. 21). The paper does not test alternate values.
  • Distance function psi(y,m)=|y-m|/(y+epsilon) = not specified (epsilon small)
    The soft-label distance is defined with this ad hoc form and an unspecified epsilon (Section 4.2, Eq. 20). The method's benefit over hard labels may depend on this choice.
  • Tree height / number of leaf nodes = 64 (7 levels)
    They use a 7-level tree (64 leaves) but Appendix D shows modest sensitivity; it is still a chosen hyperparameter.
  • Conversion delay window = 3 days
    The label is defined as conversions tracked within 3 days of impressions (Section 5.1.1). This window is an arbitrary modeling choice that affects all results.
assumptions (4)
  • domain assumption PCOC of the ranking model has a narrow value range and does not exhibit a long-tail distribution
    This is the basis of VRMP (Section 3, Section 4.3, Figure 2b). If violated, the proxy-label regression would not avoid the long-tail difficulty.
  • domain assumption Equal-frequency bucketing on integer long-tail labels puts many samples near bucket boundaries
    This motivates the label smoothing; if false, the discontinuity problem is overstated (Section 1, Section 4.2).
  • ad hoc to paper The soft-label mapping h(psi) preserves ordering and yields smooth gradients
    The specific functional forms psi(y,m) and h(x) are chosen ad hoc with no derivation (Section 4.2, Section 4.3).
  • domain assumption The aggregated pCTCVR z is a useful input feature and its prediction bias is stable
    VRMP converts PCOC back to y via z, and the central method relies on z being informative (Section 3, Section 4.3).

how reviews work

0 comments
Cite this review

Pith. "Pith review of LDACP: Long-Delayed Ad Conversions Prediction Model for Bidding Strategy." pith.science (2026). https://pith.science/paper/STWTRK6J

@misc{pith2026241116095,
  author       = {Pith},
  title        = {Pith review of: LDACP: Long-Delayed Ad Conversions Prediction Model for Bidding Strategy},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/STWTRK6J}},
  note         = {Machine review of arXiv:2411.16095}
}
read the original abstract

In online advertising, once an ad campaign is deployed, the automated bidding system dynamically adjusts the bidding strategy to optimize Cost Per Action (CPA) based on the number of ad conversions. For ads with a long conversion delay, relying solely on the real-time tracked conversion number as a signal for bidding strategy can significantly overestimate the current CPA, leading to conservative bidding strategies. Therefore, it is crucial to predict the number of long-delayed conversions. Nonetheless, it is challenging to predict ad conversion numbers through traditional regression methods due to the wide range of ad conversion numbers. Previous regression works have addressed this challenge by transforming regression problems into bucket classification problems, achieving success in various scenarios. However, specific challenges arise when predicting the number of ad conversions: 1) The integer nature of ad conversion numbers exacerbates the discontinuity issue in one-hot hard labels; 2) The long-tail distribution of ad conversion numbers complicates tail data prediction. In this paper, we propose the Long-Delayed Ad Conversions Prediction model for bidding strategy (LDACP), which consists of two sub-modules. To alleviate the issue of discontinuity in one-hot hard labels, the Bucket Classification Module with label Smoothing method (BCMS) converts one-hot hard labels into non-normalized soft labels, then fits these soft labels by minimizing classification loss and regression loss. To address the challenge of predicting tail data, the Value Regression Module with Proxy labels (VRMP) uses the prediction bias of aggregated pCTCVR as proxy labels. Finally, a Mixture of Experts (MoE) structure integrates the predictions from BCMS and VRMP to obtain the final predicted ad conversion number.

Figures

Figures reproduced from arXiv: 2411.16095 by the authors.

Figure 1
Figure 1. For ad campaigns using oCPM bidding, the auto [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The distribution characteristics of ad conversion numbers and the PCOC of the ranking model. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. The LDACP consists of two sub-modules. The BCMS [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: An instance of a TPMS. Each edge in the tree is asso [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: The CR and average value of 𝜆 on Kuai-AD test set. 𝑦ˆ𝑓 and 𝑦ˆ𝑔 each have their distinct advantages. 𝑦ˆ integrates the strengths of both to address the challenge of predicting tail data and enhance overall performance. product name, optimization objectives and the indus…
Figure 6
Figure 6. Figure 6: An ad campaign is split into a control group and [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 8
Figure 8. Figure 8: Performance of LDACP with varying numbers of [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

24 extracted references · 21 canonical work pages

  1. [1]

    Cost per action and optimized cost per mille, 2012

    Facebook. Cost per action and optimized cost per mille, 2012. URL https:// developers.facebook.com

    work page 2012

  2. [2]

    A personalized automated bidding framework for fairness-aware online advertising

    Haoqi Zhang, Lvyin Niu, Zhenzhe Zheng, Zhilin Zhang, Shan Gu, Fan Wu, Chuan Yu, Jian Xu, Guihai Chen, and Bo Zheng. A personalized automated bidding framework for fairness-aware online advertising. In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , pages 5544–5553, 2023

    work page 2023

  3. [3]

    Estimating conversion rate in display advertising from past performance data, July 31 2018

    Kuang-chih Lee, Burkay Birant Orten, Ali Dasdan, and Wentong Li. Estimating conversion rate in display advertising from past performance data, July 31 2018. US Patent 10,037,543

    work page 2018

  4. [4]

    Modeling delayed feedback in display advertising

    Olivier Chapelle. Modeling delayed feedback in display advertising. InProceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining, pages 1097–1105, 2014

    work page 2014

  5. [5]

    Estimating con- version rate in display advertising from past erformance data

    Kuang-chih Lee, Burkay Orten, Ali Dasdan, and Wentong Li. Estimating con- version rate in display advertising from past erformance data. In Proceedings of the 18th ACM SIGKDD international conference on Knowledge discovery and data mining, pages 768–776, 2012

    work page 2012

  6. [6]

    Calibrating user response predictions in online advertising

    Chao Deng, Hao Wang, Qing Tan, Jian Xu, and Kun Gai. Calibrating user response predictions in online advertising. In Machine Learning and Knowledge Discovery in Databases: Applied Data Science Track: European Conference, ECML PKDD 2020, Ghent, Belgium, September 14–18, 2020, Proceedings, Part IV , pages 208–223. Springer, 2021

    work page 2020

  7. [7]

    Posterior probability matters: Doubly-adaptive calibration for neural predictions in online advertising

    Penghui Wei, Weimin Zhang, Ruijie Hou, Jinquan Liu, Shaoguo Liu, Liang Wang, and Bo Zheng. Posterior probability matters: Doubly-adaptive calibration for neural predictions in online advertising. In Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval , pages 2645–2649, 2022

    work page 2022

  8. [8]

    MBCT: tree-based feature-aware binning for individual uncertainty calibration

    Siguang Huang, Yunli Wang, Lili Mou, Huayue Zhang, Han Zhu, Chuan Yu, and Bo Zheng. MBCT: tree-based feature-aware binning for individual uncertainty calibration. In WWW, pages 2236–2246, 2022

    work page 2022

Show all 24 references
  1. [9]

    A deep probabilistic model for customer lifetime value prediction

    Xiaojing Wang, Tianqi Liu, and Jingang Miao. A deep probabilistic model for customer lifetime value prediction. arXiv preprint arXiv:1912.07753, 2019

    1912 arXiv

  2. [10]

    Out of the box thinking: Improving customer lifetime value modelling via expert routing and game whale detection

    Shijie Zhang, Xin Yan, Xuejiao Yang, Binfeng Jia, and Shuangyang Wang. Out of the box thinking: Improving customer lifetime value modelling via expert routing and game whale detection. In Proceedings of the 32nd ACM International Conference on Information and Knowledge Managem...

    2023

  3. [11]

    Deep neural networks for youtube recommendations

    Paul Covington, Jay Adams, and Emre Sargin. Deep neural networks for youtube recommendations. In Proceedings of the 10th ACM conference on recommender systems, pages 191–198, 2016

    2016

  4. [12]

    Billion-user customer lifetime value prediction: an industrial-scale solution from kuaishou

    Kunpeng Li, Guangcui Shao, Naijun Yang, Xiao Fang, and Yang Song. Billion-user customer lifetime value prediction: an industrial-scale solution from kuaishou. In Proceedings of the 31st ACM International Conference on Information & Knowledge Management, pages 3243–3251, 2022

    2022

  5. [13]

    Cread: A classification-restoration framework with error adaptive discretization for watch time prediction in video recommender systems

    Jie Sun, Zhaoying Ding, Xiaoshuang Chen, Qi Chen, Yincheng Wang, Kaiqiao Zhan, and Ben Wang. Cread: A classification-restoration framework with error adaptive discretization for watch time prediction in video recommender systems. In Proceedings of the AAAI Conference on Artifi...

    2024

  6. [14]

    Tree based progressive regression model for watch-time prediction in short- video recommendation

    Xiao Lin, Xiaokai Chen, Linfeng Song, Jingwei Liu, Biao Li, and Peng Jiang. Tree based progressive regression model for watch-time prediction in short- video recommendation. In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , pages 4497–4506, 2023

    2023

  7. [15]

    Deconfounding duration bias in watch-time prediction for video recommendation

    Ruohan Zhan, Changhua Pei, Qiang Su, Jianfeng Wen, Xueliang Wang, Guanyu Mu, Dong Zheng, Peng Jiang, and Kun Gai. Deconfounding duration bias in watch-time prediction for video recommendation. In Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Min...

    2022

  8. [16]

    Uncertainty-aware probabilistic travel time prediction for on-demand ride-hailing at didi

    Hao Liu, Wenzhao Jiang, Shui Liu, and Xi Chen. Uncertainty-aware probabilistic travel time prediction for on-demand ride-hailing at didi. In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , pages 4516–4526, 2023

    2023

  9. [17]

    Adaptive mixtures of local experts

    Robert A Jacobs, Michael I Jordan, Steven J Nowlan, and Geoffrey E Hinton. Adaptive mixtures of local experts. Neural computation, 3(1):79–87, 1991

    1991

  10. [18]

    Deep expectation of real and apparent age from a single image without facial landmarks

    Rasmus Rothe, Radu Timofte, and Luc Van Gool. Deep expectation of real and apparent age from a single image without facial landmarks. International Journal of Computer Vision, 126(2):144–157, 2018

    2018

  11. [19]

    A constrained deep neural network for ordinal regression

    Yanzhu Liu, Adams Wai Kin Kong, and Chi Keong Goh. A constrained deep neural network for ordinal regression. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 831–839, 2018

    2018

  12. [20]

    Deep ordinal regression network for monocular depth estimation

    Huan Fu, Mingming Gong, Chaohui Wang, Kayhan Batmanghelich, and Dacheng Tao. Deep ordinal regression network for monocular depth estimation. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 2002–2011, 2018

    2002

  13. [21]

    Soft labels for ordinal regression

    Raul Diaz and Amit Marathe. Soft labels for ordinal regression. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 4738–4747, 2019

    2019

  14. [22]

    Mean-variance loss for deep age estimation from a face

    Hongyu Pan, Hu Han, Shiguang Shan, and Xilin Chen. Mean-variance loss for deep age estimation from a face. InProceedings of the IEEE conference on computer vision and pattern recognition , pages 5285–5294, 2018

    2018

  15. [23]

    Adam: A method for stochastic optimization

    Diederik P Kingma. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980, 2014

    2014 arXiv

  16. [24]

    Self-normalizing neural networks

    Günter Klambauer, Thomas Unterthiner, Andreas Mayr, and Sepp Hochreiter. Self-normalizing neural networks. Advances in neural information processing systems, 30, 2017. A Conversion Delay for Paid Objectives As shown in Table 6, we present the average delay of paid objec- tives...

    2017

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.