Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

An inverse mixed-integer optimization framework for learning interpretable models of expert decision making

T0 review · 4 major / 5 minor · reviewed 2026-08-08 · deepseek-v4-flash

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

Pith's one-line read A unified inverse-optimization framework learns both an expert's hidden cost preferences and the soft decision rules behind observed choices.

desk verdict A genuinely new framework for joint learning of costs and decision rules, but the Amazon case study rests on an unverified feasibility assumption that could undermine its headline numbers. read the letter →

arxiv 2608.05585 v1 pith:YSJRPBBV submitted 2026-08-06 math.OC

Anurag Holani , Rishabh Gupta , John Wassick , Qi Zhang This is my paper · ORCID
classification math.OC MSC 90C1190C2790B06
keywords inverseoptimizationmixed-integerprogrammingdecisionrulesinterpretablemodelsexpertmakingimitationlearninglast-mileroutingsuboptimalityloss
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

Experts often cannot articulate why they make the choices they do, but the decisions themselves are recorded. This paper proposes an inverse-optimization framework that assumes each observed decision is an approximately optimal solution to an underlying mixed-integer program whose cost coefficients and soft 'decision rules' (logical conditions the expert tends to follow) are unknown. Rather than learning only preferences or only rules, the framework learns both at once: rules are encoded as propositional-logic constraints with learnable rewards, and the inverse problem minimizes the suboptimality of the observed decisions. Across a shift-assignment problem, a production-planning problem, and a real-world Amazon delivery-route dataset, the joint model predicts held-out decisions better than models that learn pairwise costs alone. The routing case is the headline: adding two zone-clustering rules and then refining pairwise costs cuts the Amazon Score from 0.0632 to 0.0533, roughly a 16% improvement over the distance baseline.

What carries the argument

The load-bearing object is the rule-augmented forward MILP: a base mixed-integer linear program with unknown cost coefficients $\bar{c}(u)$, augmented by a set of soft decision rules $\Omega_r(\bar{x},u)$, each represented as a propositional-logic statement reformulated into mixed-integer linear inequalities (via the Raman–Grossmann reformulation) and activated by a binary variable $z_r$ with a learnable reward $d_r(u)$. The inverse problem minimizes the suboptimality loss $\Delta_i(\theta)=c(u_i;\theta)^\top x_i-\min_{\tilde{x}\in S_i} c(u_i;\theta)^\top \tilde{x}$, and the cutting-plane algorithm of Wang (2009) solves the semi-infinite program by iteratively adding the most violated alternative solutions as cuts. This machinery turns the intractable bilevel formulation into a sequence of linear master problems and MILP forward solves.

What would settle it

Take a synthetic expert with known cost weights and known decision rules, generate decisions, then run the proposed inverse algorithm on data where some observed decisions are infeasible in the hypothesized forward model (for example, a driver route that revisits a zone non-contiguously before all its stops are served). If the learned cost weights and rule rewards do not converge to values that reproduce the expert's decisions—or if the model's prediction accuracy is no better than a cost-only baseline—the central claim would be falsified.

Watch

Extended reading notes

Core claim

The central claim is that a single inverse-optimization formulation can simultaneously recover the cost parameters of a decision-maker's optimization problem and the soft logical rules that shape their choices, and that doing so yields more accurate, interpretable models than recovering costs alone. The paper develops a rule-augmented forward problem in which each candidate decision rule is a proposition over the decision variables and context, reformulated as mixed-integer linear constraints using the Raman–Grossmann framework, and is activated by a binary variable carrying a learnable reward. The inverse problem then minimizes the suboptimality loss of the observed decisions via a cutting-plane algorithm that alternates between a master problem and forward solves. In the Amazon Last Mile Routing Research Challenge, the learned rules correspond to drivers visiting zones in contiguous hierarchical clusters, and the cluster-crossing diagnostic shows that rules, not pairwise travel costs, reproduce the block structure of human routes. The paper concludes that modeling both preferences and rules is essential when rules materially shape observed decisions.

Load-bearing premise

The framework assumes every observed decision is feasible in the forward model, including the auxiliary decision-rule variables; in the routing application this requires that each driver's stop sequence, after mapping stops to zones and compressing contiguous same-zone runs, forms a Hamiltonian tour over the zones.

Editorial extensions

If this is right

  • A model trained on historical decisions can be queried for both the expert's value system and their standing heuristics, making the reasoning auditable for training novices or certifying automated systems.
  • In routing, the recovered zone-clustering rules predict that drivers serve zones contiguously in a hierarchy; a planning system that ignores this will systematically underestimate the cost of interleaving zones.
  • Learned rule rewards quantify how strongly each hypothesized heuristic influences decisions, so an analyst can prune or prioritize rules by estimated impact.
  • The framework carries over to any mixed-integer decision problem where candidate rules can be stated as propositions, not just routing or scheduling.

Reading between the lines

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

  • A fully joint estimation of rule rewards and pairwise costs, instead of the paper's two-stage Amazon pipeline, might trade off rule strength against cost detail differently; comparing the two reveals how much of the 6.7 percent gain is attributable to the staging rather than to the rules themselves.
  • The feasibility assumption means the framework silently ignores experts whose decisions cannot be represented in the forward model, such as drivers who revisit a zone non-contiguously; extending the loss to penalize infeasible observations is a natural next step.
  • The cluster-crossing diagnostic (predicted routes crossing clusters more often than observed routes) is a cheap, model-free signal for when a forward model is missing structural rules, and could serve as a stopping rule for rule-set construction in other domains.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper proposes an inverse mixed-integer optimization framework that jointly learns the cost parameters of a forward optimization problem and the rewards of soft decision rules, expressed as propositional-logic constraints reformulated into mixed-integer form. The method minimizes a suboptimality-based loss via a cutting-plane algorithm. Three case studies are presented: a synthetic shift assignment problem, a synthetic production planning problem, and the Amazon Last Mile Routing Research Challenge. The central empirical claim is that adding learned decision rules to pairwise cost learning improves out-of-sample route prediction, with Amazon Score improving from 0.0632 (distance baseline) to 0.0533, about a 15.8% reduction, and from 0.0571 for the best model without decision rules, about a 6.7% improvement (Section 4.3.8). The framework is interpretable and the Amazon case study includes a held-out test set and a structural cluster-crossing diagnostic.

Significance. If the claims hold, the paper makes a useful contribution to interpretable imitation learning by unifying inverse optimization with soft logic-based decision rules. The framework is general, the cutting-plane algorithm is standard and has a finite-convergence argument, and the production-planning experiments demonstrate a clear predictive advantage of rule augmentation in a controlled setting. The Amazon case study is valuable because it is a real operational dataset with an established evaluation metric, and the paper goes beyond aggregate scores by reporting cluster-crossing diagnostics and route-level examples. I also credit the paper for explicitly documenting the suboptimality-loss formulation and its feasibility assumptions, which makes the analysis checkable. The main weaknesses are that the Amazon evaluation relies on an unverified feasibility assumption for the compressed zone tours and on a two-stage heuristic rather than the joint formulation, and that the production-planning results contain a sign inconsistency in the rule-reward interpretation.

major comments (4)
  1. [Section 3.3 and Section 4.3.2] The suboptimality loss in Eq. (9) is only well defined if every observed decision x_i lies in the feasible set S_i, as stated in Section 3.3. In the Amazon case study, the observed zone tour is obtained by mapping stops to zones and compressing contiguous same-zone runs. If a driver leaves a zone and later returns, the compressed sequence contains that zone more than once and therefore violates the Hamiltonian-tour constraints (17b)-(17d). The paper does not report how many of the 6,112 training or 3,072 test routes contain repeated zones, nor any feasibility diagnostic. If a non-negligible fraction of routes are infeasible, the reported improvements in Section 4.3.8 (0.0632 to 0.0533) are not computed against a valid loss. Please add statistics on repeated-zone occurrences and either restrict the evaluation to representable routes or extend the forward model to allow zone revisits (e.g., by permitting additional visits through a dummy-copy construction), and re-run the analysis.
  2. [Section 4.3.5] The Amazon case study does not solve the joint inverse optimization problem (9)-(10) that is the paper's central formulation. Instead, decision-rule rewards P^d are learned first with theta^d fixed to haversine distances, and then pairwise costs theta^d are refined with P^d fixed. The paper does not quantify the gap between this two-stage heuristic and the joint solution, nor does it justify the claim in the abstract that the approach jointly learns preferences and rules on this real-world example. The comparison in Section 4.3.8 therefore evaluates a heuristic pipeline, not the proposed framework. Please either solve the joint problem on the Amazon data, or explicitly characterize the two-stage procedure as an approximation and report a cross-validated comparison between the joint and two-stage variants, including the difference in Amazon Score.
  3. [Section 4.2.4 and Section 4.2.6] There is a sign inconsistency in the production planning case study. In the rule-augmented forward model (16a), the objective is maximized and the term involving m_pj is + sum m_pj z_pj, with z_pj = 1 indicating that the rule 'at most p products' is satisfied. Under this formulation, a positive m_pj rewards satisfaction of the rule, while a negative m_pj discourages setting z_pj = 1. However, Section 4.2.6 reports learned values m_51 = -274.08, m_52 = -289.46, and m_51 = -205.90 and interprets these large negative values as a strong preference for satisfying the at-most-five rule. With the given sign convention, negative values make z_pj = 0 optimal and thus indicate that the rule is not being used. This undermines the interpretability claim of that case study. Please correct the sign convention in the formulation or in the reported values, and verify that the learned rule weights reproduce the ground-truth rule preference under the corrected model.
  4. [Section 3.2 and Section 3.3] The inverse problem in (8)-(9) defines the observation x_i as the full decision vector x := (x_bar, y, z), including the auxiliary literal variables y and the rule-indicator variables z. The case studies, however, only observe the base decisions (assignments, production quantities, or zone tours); the paper never states how y_i and z_i are assigned to each observed decision. This matters because the suboptimality loss c(u_i;theta)^T x_i - min_{x in S_i} c(u_i;theta)^T x and the cut-generation test in Algorithm 1 depend on the chosen z_i and y_i. For example, in the shift assignment and production planning models, z_i can be set to 0 or 1 depending on whether the rule is considered satisfied, and the loss value changes accordingly. Please specify the canonical completion rule used in the experiments (for instance, set y_i from the literal definitions and set z_i to the value that makes the observation feasible while minimizing or maximizing the rule-reward term), and confirm that the same rule is applied in all three case studies.
minor comments (5)
  1. [Section 4.1.3] There is a typo in the sentence introducing the hypothesized model: 'the analyst knows tat some of the workers' should read 'the analyst knows that some of the workers.'
  2. [Section 4.2.6] The reported rule weights contain a repeated index: 'm_51 = -274.08, m_52 = -289.46, and m_51 = -205.90' should presumably be m_53 = -205.90 for the third facility.
  3. [Section 4.3.8] The distance-based baseline is reported as 0.0632 in the paragraph on decision-rule depth and as 0.0633 in the later comparison against the best no-rules model; please make the numbers consistent.
  4. [Section 4.3.4 and Table 1] The cluster-rule labels in Tables 1-3 use notation such as 'W-x.y' and 'W-x.Z' that is introduced informally; a short sentence defining each component (W, x, y, Z) and the meaning of the dot would improve readability.
  5. [Algorithm 1 and Section 4.3.5] The cutting-plane algorithm is described for the joint formulation, but in the Amazon experiments it is applied to the two-stage procedure; please clarify in the text whether Algorithm 1 is run separately for P^d and theta^d or whether the two-stage approach uses a different termination criterion.

Circularity Check

0 steps flagged · score 2.0 of 10

No circularity: the inverse-optimization derivation is self-contained, the Amazon and production-planning improvements are genuine out-of-sample results, and the only self-citations are background references.

full rationale

The paper's central derivation chain is not circular. Section 3.3 defines the suboptimality loss Δ_i(θ) = c(u_i;θ)ᵀx_i − min_{x̃∈S_i} c(u_i;θ)ᵀx̃ and problem (9) minimizes this loss over θ; because the loss is evaluated against the observed decision x_i and the minimum over the feasible set S_i, it is not defined in terms of the inferred parameters beyond the cost vector itself, so minimizing it is a genuine estimation problem rather than a tautology. The cutting-plane algorithm terminates by checking whether c(u_i;θ)ᵀx_i ≤ c(u_i;θ)ᵀx* for all i, a standard optimality condition, not a construction of the answer. The synthetic case studies generate data from explicit ground-truth models (Sections 4.1.2 and 4.2.2) and then estimate parameters; the production-planning study reports relative error on separate test splits (Section 4.2.5), and the Amazon study evaluates on 3,072 held-out test routes using the externally defined Amazon Score (Sections 4.3.7–4.3.8). Thus the headline improvements (0.0632 → 0.0533 with rules; 0.0571 → 0.0533 best no-rules comparison) are out-of-sample predictions, not fitted values renamed as predictions. The self-citations to Gupta and Zhang (2022, 2023) appear in Section 3.2 only as general background for data-driven inverse optimization; the specific suboptimality-loss formulation and convergence argument are attributed to Moghaddass and Terekhov (2020, 2021) and Wang (2009), so no load-bearing conclusion rests on the authors' own prior results. The feasibility assumption x_i ∈ S_i in Section 3.3 is an empirical regularity condition; if the Amazon zone-compression procedure produces non-Hamiltonian sequences for routes that revisit zones, the suboptimality loss would be undefined for those observations, but that is a validity/correctness risk, not a circularity, because no equation reduces by definition to its own input. Overall, the paper's central claims have independent content and are supported by genuine out-of-sample evaluation.

Assumptions & free parameters 7 free parameters · 6 assumptions · 0 invented entities

The ledger mostly contains the model parameters that the method is designed to learn; they are fitted to the training data rather than derived. The key modeling assumptions are feasibility of observations and known hard constraints. No new physical entities are postulated.

free parameters (7)
  • Shift assignment time-profile polynomial coefficients a0..a4
    Learned via IO to approximate the ground-truth profile phi(t)=80 sin(2 pi t/10); estimated values shown in Figure 1 but not tabulated.
  • Shift assignment rule reward parameters gamma_n(u)
    Parameterized as -c_n for u <= ubar_n and zero otherwise, with c_n and ubar_n learned per worker from data.
  • Production planning objective weights alpha, beta, gamma = alpha=48.80, beta=1.00, gamma=1.00 (500 training instances)
    Learned via IO; alpha dominates, consistent with total return being the main objective.
  • Production planning rule weights m_{pj} = m_{51}=-274.08, m_{52}=-289.46, m_{53}=-205.90
    Learned via IO; large negative values for the at-most-five-products rule at each facility, consistent with ground truth.
  • Amazon pairwise zone costs theta^d_{ij}
    Initialized with haversine distances and refined by IO over 30-40 cutting-plane iterations.
  • Amazon clustering rule rewards P^d_k
    Learned for the selected L1 (W-x.Z) and L2 (W-x) clustering rules per depot.
  • Big-M constants in MILP reformulations and cuts
    Chosen sufficiently large to enforce logic implications; exact values not reported, which can affect numerical behavior.
assumptions (6)
  • domain assumption Every observed decision is feasible in the hypothesized forward model, i.e., x_i in S_i.
    Stated in Section 3.3; needed for the suboptimality loss to be non-negative and well defined.
  • domain assumption The feasible set S_i is non-empty and bounded for every context u_i.
    Assumed in Section 3.3 to ensure the subproblem has an optimal solution and the cutting-plane convergence argument applies.
  • domain assumption The unknown cost vector c(u;theta) is affine in theta.
    Assumed in Section 3.2; restricts the learnable class of objective functions.
  • domain assumption The hard constraints in the base FOP are known and correct.
    Assumed in Section 2.1; if a hard constraint is misspecified, learned preferences and rules may absorb the error.
  • domain assumption For the routing application, the observed zone tour is a Hamiltonian tour after compressing contiguous same-zone visits.
    Implicit in Section 4.3.2; required for the forward TSP model to represent the observed route.
  • standard math An optimal solution of the cutting-plane subproblem lies at an extreme point of conv(S_i), a bounded polyhedron with finitely many extreme points.
    Invoked in Section 3.4 following Wang (2009) to guarantee finite convergence.

how reviews work

0 comments
Cite this review

Pith. "Pith review of An inverse mixed-integer optimization framework for learning interpretable models of expert decision making." pith.science (2026). https://pith.science/paper/YSJRPBBV

@misc{pith2026260805585,
  author       = {Pith},
  title        = {Pith review of: An inverse mixed-integer optimization framework for learning interpretable models of expert decision making},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YSJRPBBV}},
  note         = {Machine review of arXiv:2608.05585}
}
read the original abstract

Understanding how experts make decisions and being able to transfer that knowledge is important, especially in complex engineering applications. It is highly valuable for training novices, improving the performance of human-machine systems, and potentially enabling fully autonomous systems that perform as well as human experts. However, an expert's decision-making strategy, developed through years of experience, is often not directly accessible, since the implicit preferences and decision rules involved can be difficult to specify explicitly. This has motivated the use of observed decisions made by the expert to learn an interpretable model that captures the expert's decision-making process. In this work, we develop an inverse optimization approach to jointly learn the decision-maker's preferences (or perceived costs) and the decision rules governing their choices. We demonstrate the general applicability of our approach using three case studies that consider a shift assignment problem, a production planning problem, and a real-world routing problem, respectively. Across these case studies, modeling both perceived costs and decision rules leads to better predictions, highlighting the value of the proposed framework and its greater flexibility in capturing and replicating expert decision making.

Figures

Figures reproduced from arXiv: 2608.05585 by the authors.

Figure 1
Figure 1. Actual vs. estimated values of ϕ(t) (a) and γ(u) (b) 14 [PITH_FULL_IMAGE:figures/full_fig_p014_1.png] view at source ↗
Figure 2
Figure 2. Prediction accuracy on synthetic production planning data with a limit of five products per facility plexity. For 500 training instances, the estimated weights were α = β = γ = 1.00. All three parameters collapsing to the same value indicates that the inverse problem effectively chooses a nearly flat objective to reduce suboptimality in the presence of severe model mismatch [PITH_FULL_IMAGE:figures/full_fig_p019_2.png] view at source ↗
Figure 3
Figure 3. compares the relative test error of the rule-augmented model and the no-rules model for varying limits on the number of products per facility. When the rule has a strong effect on the production plan (e.g., low limits on the number of products), the no-rules model exhibits large prediction errors, reflecting its inability to capture the planner’s complexity preferences. In contrast, the rule-augmented model maintain… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Amazon score as a function of the number of decision rules and training iterations. Decision rules versus pairwise cost learning.We next quantify how much improvement is achieved by learning pairwise costs θ d after learning decision-rule rewards [PITH_FULL_IMAGE:figu…
Figure 5
Figure 5. Figure 5: L1 and L2 cluster crossing as percentages of the actual route. rules, to see the impact which learning these decision rule has on the overall accuracy of our route predictions, and discuss how learning the clustering of zones helps improve the final predictions. (a) Ob…
Figure 6
Figure 6. Figure 6: Actual and predicted routes: Example 1 We first consider the route shown in [PITH_FULL_IMAGE:figures/full_fig_p026_6.png]
Figure 7
Figure 7. Figure 7: Actual and predicted routes: Example 2 To further examine the interaction between multiple decision rules, [PITH_FULL_IMAGE:figures/full_fig_p027_7.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Uncovering expert objectives in production planning via inverse optimization: An industrial case study

    math.OC 2026-08 conditional novelty 5.0 of 10

    Inverse optimization of a mixed-integer production planning model on 50 training plans reveals that Dow planners weight avoiding understock and stable cycle lengths most heavily.

Reference graph

Works this paper leans on

22 extracted references · 19 canonical work pages · cited by 1 Pith paper

  1. [8]

    2022.0382

    doi: 10.1287/opre. 2022.0382. Felipe Codevilla, Eder Santana, Antonio M L ´opez, and Adrien Gaidon. Exploring the limitations of behavior cloning for autonomous driving. InProceedings of the IEEE/CVF international confer- ence on computer vision, pages 9329–9338,

    work page arXiv 2022

  2. [11]

    Rishabh Gupta and Qi Zhang

    doi: 10.1287/ijoc.2022.1162. Rishabh Gupta and Qi Zhang. Efficient learning of decision-making models: A penalty block coordinate descent algorithm for data-driven inverse optimization.Computers & Chemical Engi- neering, 170:108123, February

    work page arXiv 2022

  3. [12]

    LLC Gurobi Optimization

    doi: 10.1016/j.compchemeng.2022.108123. LLC Gurobi Optimization. Gurobi optimizer reference manual,

    work page arXiv 2022

  4. [16]

    Mahsa Moghaddass and Daria Terekhov

    doi: 10.1287/trsc.2022.1173. Mahsa Moghaddass and Daria Terekhov. Inverse integer optimization with an imperfect observa- tion.Operations Research Letters, 48(6):763–769,

    work page arXiv 2022

  5. [18]

    Christoph Molnar

    doi: 10.1007/s10107-017-1216-6. Christoph Molnar. Interpretable Machine Learning,

    work page doi:10.1007/s10107-017-1216-6

  6. [19]

    npublisher: Nature Publishing Group

    doi: 10.1038/s42256-019-0048-x. npublisher: Nature Publishing Group. Lizhi Wang. Cutting plane algorithms for the inverse mixed integer linear programming problem. Operations Research Letters, 37(2):114–116, March

    doi:10.1038/s42256-019-0048-x

  7. [21]

    Pedro Zattoni Scroccaro, Piet Van Beek, Peyman Mohajerin Esfahani, and Bilge Atasoy

    doi: 10.1109/TCYB.2024.3395626. Pedro Zattoni Scroccaro, Piet Van Beek, Peyman Mohajerin Esfahani, and Bilge Atasoy. Inverse Optimization for Routing Problems.Transportation Science, page trsc.2023.0241, July

    work page arXiv 2024

  8. [22]

    doi: 10.1287/trsc.2023.0241. 31

    work page arXiv 2023

Show all 22 references
  1. [1992]

    doi: 10.1007/BF01585693. D. Burton and Ph. L. Toint. On the use of an inverse shortest paths algorithm for recovering linearly correlated costs.Mathematical Programming, 63(1):1–22, January

    doi

  2. [2004]

    Amazon.com Inc

    doi: 10.1145/1015330.1015430. Amazon.com Inc. Amazon Last-Mile Routing Research Challenge|supported by the MIT Center for Transportation & Logistics, 2021a. URLhttps://routingchallenge.mit.edu/. Amazon.com Inc. Amazon routing challenge: Scoring, 2021b. URLhttps://github.com/ M...

  3. [2009]

    Maryam Zare, Parham M

    doi: 10.1016/j.orl.2008.12.001. Maryam Zare, Parham M. Kebria, Abbas Khosravi, and Saeid Nahavandi. A Survey of Imitation Learning: Algorithms, Recent Developments, and Challenges.IEEE Transactions on Cybernetics, 54(12):7173–7186, December

    2008 doi

  4. [2011]

    doi: 10.1109/ISIC.2011.6045410

    ISBN 978-1-4577-1104-6. doi: 10.1109/ISIC.2011.6045410. Himabindu Lakkaraju, Stephen H. Bach, and Jure Leskovec. Interpretable Decision Sets: A Joint Framework for Description and Prediction. InProceedings of the 22nd ACM SIGKDD International 30 Conference on Knowledge Discove...

    2011

  5. [2014]

    29 Timothy C

    doi: 10.1287/opre.2014.1267. 29 Timothy C. Y. Chan, Rafid Mahmood, and Ian Yihang Zhu. Inverse Optimization: Theory and Applications.Operations Research, page opre.2022.0382, December

    2014

  6. [2015]

    Jeff Bezanson, Alan Edelman, Stefan Karpinski, and Viral B Shah

    doi: 10.1007/s10107-014-0819-4. Jeff Bezanson, Alan Edelman, Stefan Karpinski, and Viral B Shah. Julia: A fresh approach to numerical computing.SIAM Review, 59(1):65–98,

    doi

  7. [2016]

    doi: 10.1145/2939672.2939874

    ISBN 978-1-4503-4232-2. doi: 10.1145/2939672.2939874. Daniel Merch ´an, Jatin Arora, Julian Pachon, Karthik Konduri, Matthias Winkenbach, Steven Parks, and Joseph Noszek. 2021 Amazon Last Mile Routing Research Challenge: Data Set. Transportation Science, 58(1):8–11, January

    2021

  8. [2017]

    Arezou Keshavarz, Yang Wang, and Stephen Boyd

    doi: 10.1145/3054912. Arezou Keshavarz, Yang Wang, and Stephen Boyd. Imputing a convex objective function. In2011 IEEE International Symposium on Intelligent Control, pages 613–619, Denver, CO, USA, September

    doi

  9. [2018]

    Learning accurate and interpretable decision trees

    Maria-Florina Balcan and Dravyansh Sharma. Learning accurate and interpretable decision trees. arXiv preprint arXiv:2405.15911,

    arXiv

  10. [2019]

    Xiaotong Guo, Baichuan Mo, and Qingyi Wang

    doi: 10.1145/3236009. Xiaotong Guo, Baichuan Mo, and Qingyi Wang. Amazon last-mile delivery trajectory prediction using hierarchical tsp with customized cost matrix.arXiv preprint arXiv:2302.02102,

    arXiv

  11. [2021]

    Peyman Mohajerin Esfahani, Soroosh Shafieezadeh-Abadeh, Grani A

    doi: 10.1007/s11590-021-01721-4. Peyman Mohajerin Esfahani, Soroosh Shafieezadeh-Abadeh, Grani A. Hanasusanto, and Daniel Kuhn. Data-driven inverse optimization with imperfect information.Mathematical Programming, 167(1):191–234, January

    doi

  12. [2022]

    doi: 10.1287/ijoc.2021.1138. D. Burton and Ph. L. Toint. On an instance of the inverse shortest paths problem.Mathematical Programming, 53(1):45–61, January

    2021

  13. [2023]

    Dimitris Bertsimas, Vishal Gupta, and Ioannis Ch

    doi: 10.1016/j.ajodo.2023.09.011. Dimitris Bertsimas, Vishal Gupta, and Ioannis Ch. Paschalidis. Data-driven estimation in equilib- rium using inverse optimization.Mathematical Programming, 153(2):595–633, November

    2023 doi

  14. [2024]

    Barry de Ville

    doi: 10.1287/trsc.2022.1185. Barry de Ville. Decision trees.WIREs Computational Statistics, 5(6):448–455,

    2022

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.