Aligning Learning and Endogenous Decision-Making

Brian Quanz; Georgia Perakis; Pavithra Harsha; Rares Cristian

REVIEW 4 major objections 5 minor 30 references

Aligning Learning and Endogenous Decision-Making

T0 review · 4 major / 5 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 This paper argues that decision-focused learning under endogenous uncertainty should minimize squared error in decision cost rather than prediction error, and that a robust min-max variant over cost-consistent models yields near-optimal…

desk verdict The core task-loss idea is a real contribution, but Theorem 2's proof assumes what it needs to prove, so the paper needs a corrected guarantee before it is publishable. read the letter →

arxiv 2507.00851 v1 pith:UGNHEU2M submitted 2025-07-01 cs.LG

Rares Cristian , Pavithra Harsha , Georgia Perakis , Brian Quanz This is my paper · ORCID

classification cs.LG

keywords endogenousuncertaintydecision-focusedlearningtask-basedlossrobustoptimizationend-to-endinformationgatheringpricingassortment

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

Decisions often change the very quantities a model must predict: prices alter demand, stock levels alter sales, and you only ever observe the outcome of the action you actually took. This paper claims that the right training objective in such settings is a task-based loss, the squared difference between the cost of a historical decision under the model's forecast and the cost that was actually observed. On top of that loss, the paper builds an uncertainty set of all models whose task-based error on the data is small, and chooses decisions that maximize the worst-case predicted cost over that set; it proves this robust decision is near-optimal in expected cost with high probability and that the needed uncertainty radius shrinks as data grows. If these claims hold, they give a practical, principled replacement for counterfactual information in pricing, assortment, and other decision-focused learning problems.

What carries the argument

The load-bearing object is the cost-calibrated forecast: a function $f^*$ for which $c(v,f^*(x,v))=\mathbb{E}[c(v,z)\mid v,x]$, so that plugging a point forecast into the cost function reproduces expected cost. Training uses the task-based loss $E(f)$ as an empirical surrogate for this identity, and robustness is carried by the uncertainty set $U_\varepsilon=\{f\in F: E(f)\le \beta+\varepsilon\}$, from which the decision rule $v^*_{\mathrm{robust}}=\arg\min_v \max_{f\in U_\varepsilon} c(v,f(x,v))$ is formed. The theoretical arguments combine Rademacher complexity bounds for the generalization gap with covering numbers, Lipschitz continuity of the models in their parameters, and a martingale concentration inequality for the robust coverage guarantee. Computationally, the paper contributes an exact mixed-integer reformulation when costs are piecewise-linear convex, a sampling-based training heuristic, and a cutting-plane algorithm for the outer minimax problem.

What would settle it

Use a nonlinear cost such as $c(v,z)=z^2$ with bounded $z$ and a hypothesis class too narrow to contain any cost-calibrated predictor; simulate many datasets, compute the population minimizer $f^*$, and record how often $f^*$ actually falls in $U_{\varepsilon_N}$ at the claimed confidence. If the empirical coverage is below $1-\delta$, the theorem's coverage claim as stated fails for that setting.

Watch

Extended reading notes

Core claim

The central claim is that aligning learning with downstream decisions under endogenous uncertainty reduces to minimizing $E(f)=\sum_n (c(v_n,f(x_n,v_n))-c(v_n,z_n))^2$, and that this objective is the correct substitute for the unavailable counterfactual data. The paper shows under a cost-calibration condition, $c(v,f^*(x,v))=\mathbb{E}[c(v,z)\mid v,x]$, the forecast $f^*$ induces optimal decisions. Its robust theorem then states that for the uncertainty set $U_\varepsilon=\{f\in F: E(f)\le \beta+\varepsilon\}$ with $\varepsilon$ chosen according to covering numbers and Lipschitz constants, the best-in-class predictor lies in $U_\varepsilon$ with probability $1-\delta$, so the min-max rule $v^*_{\mathrm{robust}}=\arg\min_v \max_{f\in U_\varepsilon} c(v,f(x,v))$ attains near-optimal expected cost; for linear classes $\varepsilon=O((d\log N+\log 1/\delta)/N)$. The paper also extends the framework to two-stage information-gathering problems, where the first decision is which random variable to observe before a second-stage forecast and decision.

Load-bearing premise

The robust guarantee assumes the best model inside the chosen class is exactly cost-calibrated, meaning its predicted cost equals the true conditional expected cost at every action, and for nonlinear cost functions, being the minimizer of squared cost error does not by itself force that property.

Editorial extensions

If this is right

Under endogenous uncertainty, the task-based loss replaces unobservable counterfactuals with observed decision costs, so no action-outcome pairs beyond history are needed to train a decision-aware model.
The robust uncertainty set contains the true best-in-class predictor with probability at least $1-\delta$, and its radius shrinks as $O((d\log N+\log 1/\delta)/N)$ for linear models, so protection against model error becomes cheaper with more data.
Decisions obtained by the min-max rule inherit near-optimal expected cost with high probability, which is the property a planner needs before deploying learned forecasts in pricing or inventory.
The same end-to-end machinery applies to two-stage information-gathering problems, where the first decision selects which random variable to observe and the second decision is made from the updated forecast.
On the paper's pricing, assortment, and electricity-scheduling experiments, the approach consistently beats two-stage prediction, direct reward learning, and robust offline baselines in achieved cost.

Reading between the lines

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

Because robustness is defined over models rather than over outcome realizations, the uncertainty set can be composed with exogenous end-to-end losses for mixed problems, not just the polling example in the paper.
The cost-calibration diagnosis suggests a testable extension: augment the uncertainty-set radius with an approximation-error term when no function in the class is exactly cost-calibrated, and re-check the coverage guarantee.
The information-gathering formulation generalizes to active feature acquisition in supervised learning, where the first-stage choice is which covariate to buy information about before the forecast and decision.
The assortment finding that the optimal predictor can be a quantile of demand hints that task-based losses may automatically learn risk-averse statistics for nonlinear costs, which could be tested on other cost functions.

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

Short version: this is a legitimate attempt at a hard problem, and the core idea—training on the task loss in the endogenous setting—is worth taking seriously. The main theoretical guarantee in Theorem 2, however, is not proven as stated.

What the paper does well: the setup is honest about the absence of counterfactuals. The task-based loss (3) is a natural adaptation of exogenous end-to-end learning, but the adaptation is not trivial: you cannot evaluate decisions you did not take. The robust uncertainty set over models (13) is a sensible construction, and the two-stage information-gathering problem in Section 4 is a genuine extension. The algorithms (MIO, sampling, cutting-plane) are useful, and the experiments cover three distinct problem classes with reasonable baselines. There's no smoke and mirrors in the empirical work; the missing error bars and unreleased code are the main complaints there.

The soft spot is load-bearing. In the proof of Theorem 2 (Appendix B), the authors use the identity E[c(v_i,z_i)|v_i,x_i] = c(v_i,f*(v_i,x_i)) 'by definition.' But f* is only the population minimizer of squared cost error. For nonlinear cost functions and restricted model classes, the minimizer need not be cost-calibrated. The cross term in Eq. (78) then does not vanish, and the martingale bound leading to coverage can fail. So the high-probability statement that f* lies in U_epsilon is not established in the stated generality. The abstract's stronger claim—that the robust decision achieves near-optimal expected cost—would need a separate argument bounding decision cost in terms of epsilon; the paper only gives containment of f*. This is not a minor typo; it is the central guarantee.

Other issues are minor by comparison: Theorem 1 has a sum/average typo, the choice of epsilon is left to the user, and the experiments lack error bars. None of these would be fatal if the proof were fixed.

Who this is for: researchers in decision-focused learning and contextual optimization. They will find the formulation and the information-gathering extension useful even while the theory is under repair. My advice: send this to a serious referee rather than desk-reject. It deserves a chance to be repaired. If the authors can prove a correct version under stated assumptions—either by assuming a cost-calibrated function lies in F or by softening the theorem—it would be a useful contribution.

Referee Report

4 major / 5 minor

Summary. The paper proposes an end-to-end learning framework for stochastic optimization with endogenous uncertainty, in which the learning objective minimizes the squared difference between predicted task costs and observed task costs rather than predicting the uncertain quantity itself. It then introduces a robust variant that constructs an uncertainty set of models with low task loss and makes decisions by minimizing worst-case predicted cost. The authors claim a high-probability guarantee that the true optimal predictor lies in this uncertainty set, and the abstract claims that the robust approach captures near-optimal decisions as a function of data. The paper also extends the framework to a two-stage information-gathering problem and presents computational experiments on pricing, assortment optimization, and electricity scheduling.

Significance. If the theoretical guarantees were valid, the paper would make a useful contribution to decision-focused learning under decision-dependent uncertainty: the problem is well motivated, the task-based loss is a natural alternative to two-stage prediction, and the proposed exact MIQP and sampling algorithms give practical paths to optimization. The extension to information-gathering is novel and the computational study is broad, covering pricing, assortment, and a real-data electricity scheduling problem. However, the central theoretical result, Theorem 2, is not established as stated because the proof relies on an unverified cost-calibration identity, and the abstract's near-optimal-decision claim is not supported by any theorem. These are load-bearing gaps, so the paper is not acceptable in its current form.

major comments (4)

[Appendix B, Eqs. (76)-(78); Theorem 2, Eq. (18)] The proof of Theorem 2 invokes the identity E[c(v_i,z_i)|v_i,x_i] = c(v_i,f*(v_i,x_i)) 'by definition' (the manuscript writes E[c(v_i,z_i)] = c(v_i,f*(v_i,z_i)) immediately before Eq. (76)). This does not follow from the definition of f* as the population minimizer of the task loss in Eq. (18); it is exactly the cost-calibration property in Eq. (2), which Theorem 2 does not assume and which need not hold when c is nonlinear in z and the hypothesis class F is restricted. Consequently the simplification in Eq. (78) is invalid: E[V_i(f)] is not equal to E[(c(v_i,f)-c(v_i,f*))^2] because the cross term containing c(v_i,f*) - E[c(v_i,z_i)|v_i,x_i] does not vanish. The variance bound in Eq. (74) and the subsequent martingale concentration therefore do not establish the claimed coverage guarantee.
[Section 2.2, Theorem 1, Eq. (11)] The statement of Theorem 1 defines the empirical loss \hat l(f) as a sum over N terms rather than an average, and the additive term is printed as (8 log(2/\delta)/N)^2 instead of a square-root term. With the sum definition, the right-hand side grows with N and the bound is not the claimed 1/sqrt(N) generalization gap; the proof in Appendix B uses an averaged loss and a square-root term. The theorem statement must be corrected before the generalization result can be evaluated.
[Section 3, Eq. (13); Theorem 2, Eq. (19); Appendix B] The uncertainty set U_epsilon in Eq. (13) and the theorem's epsilon_N in Eq. (19) are stated in terms of the summed task loss E(f) from Definition 1, but the proof of Theorem 2 in Appendix B works with the averaged loss \hat E(f). In addition, the proof concludes with an epsilon of 8 log(|N_gamma|/\delta)/t + 3M\gamma and never uses the Delta term that appears in Eq. (19). The statement and proof therefore do not match, and the stated rate for epsilon_N is not derived. The proof also drops the Lipschitz constant L of c in the bound in Eq. (57).
[Abstract; Section 1.2; Section 3] The abstract and the contribution list claim that the robust approach captures near-optimal decisions with high probability, but no theorem in the paper establishes this. Theorem 2 only proves that f* belongs to the uncertainty set U_epsilon; it does not bound E[c(v^R,z)] - min_{v in V} E[c(v,z)] for the robust decision v^R in terms of epsilon or the sample size N. Without such a bound, the headline guarantee that the robust decision is near-optimal is unsupported.

minor comments (5)

[Table 1] The row for 'Predict then optimize' reports an average difference of 710%, which appears to be a typo (possibly 71.0%); please verify the value.
[Proof of Corollary 1, Appendix B] The proof says one can choose gamma = 1/log N, but this yields a 3M/log N term in epsilon, not the O(M/N) rate claimed in the corollary; choosing gamma of order 1/N would give the stated rate.
[Eq. (76)] The expression 'c(v_i, f*(v_i, z_i))' should read 'c(v_i, f*(v_i, x_i))'.
[Section 5.1] The text contains the typo 'rboust' in the discussion of Figure 3; it should be 'robust'.
[Section 5] The experimental results report point estimates of average cost but no standard errors, confidence intervals, or repeated-seed variability; adding these would strengthen the comparison, particularly for the claims of consistent improvement.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the task-based loss, uncertainty-set coverage guarantee, and generalization bounds are self-contained statistical arguments; the caveats are an unstated calibratability assumption in Theorem 2's proof (a soundness gap, not a circular reduction), an abstract over-claim of near-optimal decisions, and one minor non-load-bearing self-citation.

full rationale

The paper's derivation chain is self-contained. The task-based loss E(f) in (3)-(4) is the objective actually minimized; the uncertainty set (13), U_eps = {f in F : E(f) <= beta + eps}, is an empirical-loss ball around the empirical minimizer; Theorem 1 is a standard Rademacher complexity bound; Theorem 2 is a martingale concentration argument showing that the population loss minimizer f* (defined in Eq. 18) lies in U_eps with high probability, with eps_N shrinking at rate O(1/N). No fitted parameter is renamed as a prediction: the experiments (Sections 5.1-5.3) evaluate decisions on out-of-sample features against ground-truth realizations, not against quantities used to fit the models. Two caveats, neither circular. First, the proof of Theorem 2 (Appendix B) states, between Eqs. (76) and (77): 'Recall by definition that E[c(v_i, z_i)] = c(v_i, f*(v_i, x_i))' (the paper's text has a typo, writing f*(v_i, z_i), and suppresses the conditioning on v_i, x_i). But f* is defined as the population minimizer of E[(c(v, f(x, v)) - c(v, z))^2]; pointwise calibration c(v, f*(x, v)) = E[c(v, z) | x, v] follows from that definition only if F contains a cost-calibrated function (e.g., an unrestricted class with invertible c), an expressiveness condition the theorem never states. For restricted linear or neural classes the identity can fail, so E[V_i(f)] = E[(c(v_i, f) - c(v_i, f*))^2] (Eq. 78) and the variance bound (74) can break; the coverage guarantee is therefore not established as stated. This is a missing-assumption soundness gap, not an equation reducing to its own input by construction. Second, the abstract claims the robust approach 'can capture near-optimal decisions with high probability as a function of data,' but Theorem 2 proves only containment f* in U_eps; no bound on E[c(v^R, z)] - min_v E[c(v, z)] in terms of eps is supplied, so the near-optimality assertion is unsupported. The only author-overlap citation, Cristian et al. (2023), appears in Related Work and in Algorithm 4 as one option for computing dv*/dz; neither the theorems nor the main conclusions depend on it, so it is not load-bearing. Proposition 1 is a correct continuity argument for the existence of a calibrated point forecast. Verdict: no significant circularity; score 2 reflects one minor, non-load-bearing self-citation alongside non-circular proof gaps.

Assumptions & free parameters 1 free parameters · 5 assumptions · 0 invented entities

The central method rests on a standard data-generation assumption, a Lipschitz smoothness assumption, and an unstated cost-calibration assumption that appears in the proof of Theorem 2. The robustness parameter epsilon is a free hyperparameter in practice.

free parameters (1)

epsilon (uncertainty set size) = tuned per experiment, e.g., 0.2 in Figure 4
Defines U_epsilon in Eq. (13). The theory prescribes a value epsilon_N depending on unknown Delta, M, gamma, so in practice epsilon is chosen by hand. Reported performance depends on this choice.

assumptions (5)

domain assumption Assumption 1: decisions v_i are generated by an algorithm that depends only on history H_{i-1}
This data-generation process rules out decisions that depend on unobserved confounders or on current z_i; used throughout the learning formulation in Section 2.
domain assumption Assumption 2: the hypothesis class f_theta is M-Lipschitz with respect to theta (Eq. 17)
Needed for the covering-number argument in Theorem 2 and Corollary 1.
ad hoc to paper The population minimizer f* of the task-based loss satisfies c(v, f*(x,v)) = E[c(v,z)|v,x] 'by definition'
Used in Appendix B proof of Theorem 2 (equations 75-78). Not implied by f* being the argmin of E[(c(v,f(x,v))-c(v,z))^2] for nonlinear c; this unstated cost-calibration assumption is load-bearing for the martingale variance bound.
standard math Intermediate value theorem to show existence of z_hat in convex hull with c(v,z_hat) = E[c(v,z)] (Proposition 1)
Standard mathematical result used to assert the existence of a point forecast matching expected cost at a given decision.
standard math Rademacher complexity bounds and martingale concentration inequalities (Bartlett and Mendelson 2002, Lemma 1)
Standard tools used to prove Theorem 1 and Theorem 2.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Aligning Learning and Endogenous Decision-Making." pith.science (2026). https://pith.science/paper/UGNHEU2M

@misc{pith2026250700851,
  author       = {Pith},
  title        = {Pith review of: Aligning Learning and Endogenous Decision-Making},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UGNHEU2M}},
  note         = {Machine review of arXiv:2507.00851}
}

read the original abstract

Many of the observations we make are biased by our decisions. For instance, the demand of items is impacted by the prices set, and online checkout choices are influenced by the assortments presented. The challenge in decision-making under this setting is the lack of counterfactual information, and the need to learn it instead. We introduce an end-to-end method under endogenous uncertainty to train ML models to be aware of their downstream, enabling their effective use in the decision-making stage. We further introduce a robust optimization variant that accounts for uncertainty in ML models -- specifically by constructing uncertainty sets over the space of ML models and optimizing actions to protect against worst-case predictions. We prove guarantees that this robust approach can capture near-optimal decisions with high probability as a function of data. Besides this, we also introduce a new class of two-stage stochastic optimization problems to the end-to-end learning framework that can now be addressed through our framework. Here, the first stage is an information-gathering problem to decide which random variable to poll and gain information about before making a second-stage decision based off of it. We present several computational experiments for pricing and inventory assortment/recommendation problems. We compare against existing methods in online learning/bandits/offline reinforcement learning and show our approach has consistent improved performance over these. Just as in the endogenous setting, the model's prediction also depends on the first-stage decision made. While this decision does not affect the random variable in this setting, it does affect the correct point forecast that should be made.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 23 canonical work pages

[1]

, " * write output.state after.block = add.period write newline

ENTRY address author booktitle chapter doi edition editor eid howpublished institution isbn issn journal key month note number organization pages publisher school series title type url volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.all := #1...
[2]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in "" FUNCTION format.date year ...
[3]

Advances in neural information processing systems 32

Agrawal A, Amos B, Barratt S, Boyd S, Diamond S, Kolter JZ (2019 a ) Differentiable convex optimization layers. Advances in neural information processing systems 32

work page 2019
[4]

Operations Research 67(5):1453--1485

Agrawal S, Avadhanula V, Goyal V, Zeevi A (2019 b ) Mnl-bandit: A dynamic learning approach to assortment selection. Operations Research 67(5):1453--1485

work page 2019
[5]

Assortment Optimization under the Decision Forest Model

Akchen YC, Misic VV (2021) Assortment optimization under the decision forest model. arXiv preprint arXiv:2103.14067

work page Pith review arXiv 2021
[6]

International conference on machine learning, 136--145 (PMLR)

Amos B, Kolter JZ (2017) Optnet: Differentiable optimization as a layer in neural networks. International conference on machine learning, 136--145 (PMLR)

work page 2017
[7]

Advances in Neural Information Processing Systems 34:9816--9827

Arumugam D, Van Roy B (2021) The value of information when deciding what to learn. Advances in Neural Information Processing Systems 34:9816--9827

work page 2021
[8]

Journal of Machine Learning Research 3(Nov):463--482

Bartlett PL, Mendelson S (2002) Rademacher and gaussian complexities: Risk bounds and structural results. Journal of Machine Learning Research 3(Nov):463--482

2002

Show all 30 references

[9]

Advances in neural information processing systems 33:9508--9519

Berthet Q, Blondel M, Teboul O, Cuturi M, Vert JP, Bach F (2020) Learning with differentiable pertubed optimizers. Advances in neural information processing systems 33:9508--9519

2020
[10]

Operations Research 70(1):454--471

Bertsimas D, Koduri N (2022) Data-driven optimization: A reproducing kernel hilbert space approach. Operations Research 70(1):454--471

2022
[11]

Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics, 208--214 (JMLR Workshop and Conference Proceedings)

Chu W, Li L, Reyzin L, Schapire R (2011) Contextual bandits with linear payoff functions. Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics, 208--214 (JMLR Workshop and Conference Proceedings)

2011
[12]

Channel strategies and marketing mix in a connected world, 71--97 (Springer)

Cohen MC, Perakis G (2019) Optimizing promotions for multiple items in supermarkets. Channel strategies and marketing mix in a connected world, 71--97 (Springer)

2019
[13]

Proceedings of the AAAI Conference on Artificial Intelligence, volume 37, 7253--7260

Cristian R, Harsha P, Perakis G, Quanz BL, Spantidakis I (2023) End-to-end learning for optimization via constraint-enforcing approximators. Proceedings of the AAAI Conference on Artificial Intelligence, volume 37, 7253--7260

2023
[14]

Advances in neural information processing systems 30

Donti P, Amos B, Kolter JZ (2017) Task-based end-to-end model learning in stochastic optimization. Advances in neural information processing systems 30

2017
[15]

predict, then optimize

Elmachtoub AN, Grigas P (2022) Smart “predict, then optimize”. Management Science 68(1):9--26

2022
[16]

Manufacturing & Service Operations Management 22(3):461--480

Ettl M, Harsha P, Papush A, Perakis G (2020) A data-driven approach to personalized bundle pricing and recommendation. Manufacturing & Service Operations Management 22(3):461--480

2020
[17]

Proceedings of the AAAI Conference on Artificial Intelligence, volume 34, 1504--1511

Ferber A, Wilder B, Dilkina B, Tambe M (2020) Mipaal: Mixed integer program as a layer. Proceedings of the AAAI Conference on Artificial Intelligence, volume 34, 1504--1511

2020
[18]

International Conference on Machine Learning, 3199--3210 (PMLR)

Foster D, Rakhlin A (2020) Beyond ucb: Optimal and efficient contextual bandits with regression oracles. International Conference on Machine Learning, 3199--3210 (PMLR)

2020
[19]

arXiv preprint arXiv:2112.13487

Foster DJ, Kakade SM, Qian J, Rakhlin A (2021) The statistical complexity of interactive decision making. arXiv preprint arXiv:2112.13487

2021 arXiv
[20]

International conference on machine learning, 2052--2062 (PMLR)

Fujimoto S, Meger D, Precup D (2019) Off-policy deep reinforcement learning without exploration. International conference on machine learning, 2052--2062 (PMLR)

2019
[21]

IEEE Transactions on systems science and cybernetics 2(1):22--26

Howard RA (1966) Information value theory. IEEE Transactions on systems science and cybernetics 2(1):22--26

1966
[22]

Operations Research 71(2):665--687

Kash IA, Key PB, Zoumpoulis SI (2023) Optimal pricing and introduction timing of technology upgrades in subscription-based services. Operations Research 71(2):665--687

2023
[23]

Journal of the society for Industrial and Applied Mathematics 8(4):703--712

Kelley JE Jr (1960) The cutting-plane method for solving convex programs. Journal of the society for Industrial and Applied Mathematics 8(4):703--712

1960
[24]

International Conference on Artificial Intelligence and Statistics, 2305--2313 (PMLR)

Kim W, Iyengar G, Zeevi A (2024) A doubly robust approach to sparse reinforcement learning. International Conference on Artificial Intelligence and Statistics, 2305--2313 (PMLR)

2024
[25]

Advances in Neural Information Processing Systems 32

Kumar A, Fu J, Soh M, Tucker G, Levine S (2019) Stabilizing off-policy q-learning via bootstrapping error reduction. Advances in Neural Information Processing Systems 32

2019
[26]

Advances in Neural Information Processing Systems 33:1179--1191

Kumar A, Zhou A, Tucker G, Levine S (2020) Conservative q-learning for offline reinforcement learning. Advances in Neural Information Processing Systems 33:1179--1191

2020
[27]

(2020) Smart predict-and-optimize for hard combinatorial optimization problems

Mandi J, Stuckey PJ, Guns T, et al. (2020) Smart predict-and-optimize for hard combinatorial optimization problems. Proceedings of the AAAI Conference on Artificial Intelligence, volume 34, 1603--1610

2020
[28]

arXiv preprint arXiv:2306.10374

Sadana U, Chenreddy A, Delage E, Forel A, Frejinger E, Vidal T (2023) A survey of contextual optimization methods for decision making under uncertainty. arXiv preprint arXiv:2306.10374

2023
[29]

arXiv preprint arXiv:1912.02175

Vlastelica M, Paulus A, Musil V, Martius G, Rol \' nek M (2019) Differentiation of blackbox combinatorial solvers. arXiv preprint arXiv:1912.02175

2019 arXiv
[30]

Advances in neural information processing systems 34:6683--6694

Xie T, Cheng CA, Jiang N, Mineiro P, Agarwal A (2021) Bellman-consistent pessimism for offline reinforcement learning. Advances in neural information processing systems 34:6683--6694

2021

Pith tools

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

[1] [1]

, " * write output.state after.block = add.period write newline

ENTRY address author booktitle chapter doi edition editor eid howpublished institution isbn issn journal key month note number organization pages publisher school series title type url volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.all := #1...

[2] [2]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in "" FUNCTION format.date year ...

[3] [3]

Advances in neural information processing systems 32

Agrawal A, Amos B, Barratt S, Boyd S, Diamond S, Kolter JZ (2019 a ) Differentiable convex optimization layers. Advances in neural information processing systems 32

work page 2019

[4] [4]

Operations Research 67(5):1453--1485

Agrawal S, Avadhanula V, Goyal V, Zeevi A (2019 b ) Mnl-bandit: A dynamic learning approach to assortment selection. Operations Research 67(5):1453--1485

work page 2019

[5] [5]

Assortment Optimization under the Decision Forest Model

Akchen YC, Misic VV (2021) Assortment optimization under the decision forest model. arXiv preprint arXiv:2103.14067

work page Pith review arXiv 2021

[6] [6]

International conference on machine learning, 136--145 (PMLR)

Amos B, Kolter JZ (2017) Optnet: Differentiable optimization as a layer in neural networks. International conference on machine learning, 136--145 (PMLR)

work page 2017

[7] [7]

Advances in Neural Information Processing Systems 34:9816--9827

Arumugam D, Van Roy B (2021) The value of information when deciding what to learn. Advances in Neural Information Processing Systems 34:9816--9827

work page 2021

[8] [8]

Journal of Machine Learning Research 3(Nov):463--482

Bartlett PL, Mendelson S (2002) Rademacher and gaussian complexities: Risk bounds and structural results. Journal of Machine Learning Research 3(Nov):463--482

2002