A Graph Theoretic Approach to Non-Anticipativity Constraint Generation in Multistage Stochastic Programs with Incomplete Scenario Sets

Alexander Vinel; Brianna Christian; Selen Cremaschi; Zuo Zheng

REVIEW 2 major objections 5 minor 34 references

A Graph Theoretic Approach to Non-Anticipativity Constraint Generation in Multistage Stochastic Programs with Incomplete Scenario Sets

T0 review · 2 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read This paper proves that a greedy spanning-tree algorithm produces a minimum-cardinality set of non-anticipativity constraints for multistage stochastic programs with gradual uncertainty and incomplete scenario sets.

desk verdict Useful NAC-reduction algorithm for gradual uncertainty on incomplete scenario sets, but the minimum-cardinality theorem is not proved and the greedy choice can matter. read the letter →

arxiv 1908.01792 v1 pith:IAECNQVY submitted 2019-08-05 math.OC

Brianna Christian , Alexander Vinel , Zuo Zheng , Selen Cremaschi This is my paper · ORCID

classification math.OC MSC 90C1590C11

keywords multistagestochasticprogrammingnon-anticipativityconstraintsendogenousuncertaintygradualrealizationincompletescenariosetsgraphconnectivityspanningtreeminimumcardinality

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

This paper establishes that the problem of generating a minimum-cardinality set of non-anticipativity constraints (NACs) for a multistage stochastic program can be solved exactly by a greedy spanning-tree algorithm, even when uncertainty is revealed gradually and the scenario set is only a subset of all possible outcomes. In such models, NACs are equality constraints that stop decisions from using information that has not yet been realized; too many of them bloat the model. The paper proves that a set of NACs is sufficient exactly when every group of scenarios that are indistinguishable at some information state is connected in a graph whose edges are scenario pairs. SNAC builds these connections from the finest information partition to the coarsest and proves the result has the smallest possible number of scenario pairs, scaling polynomially in the scenario count when the number of uncertain parameters is fixed. A computational study on pharmaceutical and manufacturing planning instances reports the same optimal objective values with far fewer constraints and lower solver times.

What carries the argument

The central object is the scenario graph $G=(S,E)$ whose vertices are scenarios and whose edges are NAC pairs, together with the family of partitions $\Pi(c)$ induced by permissible event sets $c\in C$, called cuts. Lemma 2 converts non-anticipativity into a connectivity condition: a NAC set is sufficient exactly when every block $S^c_j$ is connected in the graph restricted to that block. SNAC works from the finest cuts (largest $|c|$) down to the coarsest, adding a spanning tree within each block where needed; the downward order and the necessity argument in Definition 1 ensure every added edge is necessary, so no edge can be removed without breaking connectivity, hence the set is minimum.

What would settle it

Take a small incomplete scenario set with two information states having the same number of realized events whose scenario groups are not nested; run SNAC under all tie-breaking orders of these two states and compare the number of scenario pairs in the output. If any order yields strictly more pairs than another, Theorem 3's guarantee of a minimum-cardinality NAC set fails for that instance.

Watch

Extended reading notes

Core claim

The central claim is Theorem 3: Algorithm 1 terminates with a subset of edges that corresponds to a set of NACs enforced on the minimum number of pairs of scenarios, with an implementation running in $O(|C||S|^3)$ time as the scenario set $S$ and event-set family $C$ grow. The proof rests on two structural facts. First, for every permissible set of realized events $c\in C$, the partition $\Pi(c)$ divides scenarios into groups $S^c_j$ that are still indistinguishable, and a NAC set is sufficient iff each such group is connected in the induced subgraph of $G=(S,E)$ (Lemma 2). Second, because a coarser information state is refined by later states (Lemma 1), any edge needed to connect a group at a coarser state cannot be implied by edges added at finer states; SNAC's downward sweep over $|c|$ therefore adds only necessary edges, and the spanning-tree construction keeps each added edge necessary. The paper generalizes prior minimum-NAC-set results, which required instantaneous uncertainty realization or the full Cartesian-product scenario set, to gradual realizations with arbitrary scenario subsets.

Load-bearing premise

The proof that the greedy order yields the fewest constraints rests on the claim that, when several information states have the same number of realized events, the choice of order in which they are processed does not change how many connections are needed at coarser states later; for non-nested equal-size states this claim is assumed rather than proved.

Editorial extensions

If this is right

A user can compute the minimal NAC pair count once for a given scenario structure and reuse it across instances that differ only in objective coefficients, since SNAC reads only the uncertainty structure, not costs or revenues.
The sufficiency condition in Lemma 2 gives a cheap certificate: to check whether any proposed NAC set is sufficient, verify that every information-state group is connected in the graph; no optimization solve is needed.
For fixed numbers of uncertain parameters and events, the running-time bound $O(|C||S|^3)$ means the method remains practical as the scenario sample grows, which is the typical scaling regime for incomplete scenario sets.
In the reported case studies, SNAC preserved the optimal objective value in all instances while reducing scenario-pair counts sharply; for example, at 1,024 scenarios the full model's 523,776 pairs fell to 3,840, and total solution time dropped by roughly 30 percent for that instance.

Reading between the lines

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

Editorial inference: because Lemma 2 reduces NAC sufficiency to a connectivity check, any heuristic that removes NACs can be audited by testing each block's connectivity; the paper does not present this as a standalone verification routine, but it follows directly.
Editorial inference: the same graph and partition construction should apply to any nested family of information states, not only those derived from Cartesian products of independent uncertain parameters; SNAC's correctness depends on Lemma 1's nesting, not on the product structure.
Editorial inference: the minimal scenario-pair set produced by SNAC is a natural set to dualize in Lagrangean or progressive-hedging decompositions, since those methods dualize NACs and benefit from fewer linking constraints; the paper's experiments only use SNAC to build and solve the full model.
Editorial inference: if the equal-cardinality order assumption in Theorem 3 is false for some incomplete scenario set, the algorithm would still output a sufficient NAC set, but not necessarily a minimum one; a counterexample would appear as an order-dependent edge count in a small exhaustive search.

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.

Desk Editor's Note

read the letter

Bottom line: this paper fills a real gap—nobody else, as far as I can tell, has given a NAC-reduction algorithm for multistage stochastic programs with gradual realization and incomplete scenario sets. The graph/connectivity view is clean, and SNAC as a constructive heuristic works: it always returns a feasible NAC set, and the experiments show big reductions against the full pairwise formulation. But the headline claim—minimum cardinality—does not hold up. Theorem 3's proof is an inclusion-minimality argument, not a cardinality-minimality one. 'Necessary' is defined relative to edges added so far; an edge can be non-redundant at the time it is added and still be globally redundant because a crossing cut at the same level provides the same connectivity. The paper's own example has crossing equal-cardinality cuts ({e^1_P1} and {e^1_P2}), so this is not an exotic case. The sentence about spanning-tree choices having no effect on future iterations is asserted, not proved, and looks false in general: with the parallel treatment in Remark 1, different spanning-tree choices can give different total edge counts. You can build four-scenario examples where the algorithm outputs four edges while three suffice. The computational study does not test optimality—it only compares against the full set of pairwise constraints, so it cannot detect the gap.

What the paper does well: Lemma 2 is the right formalization; the manufacturing example is actually useful; the empirical savings in constraint count and solve time are believable; and the point that SNAC depends only on scenario structure, so it can be computed once for a family of instances, is a good practical observation. The literature review is fair and does not overclaim novelty—the new part is the gradual/incomplete case.

Recommendation: this deserves serious refereeing, because the problem is real and the algorithm is valuable even without the optimality guarantee. But the paper should not be accepted as is. The authors need to either supply a real proof of minimality (an exchange argument or lower bound independent of tie-breaking) or soften the claim to a heuristic that produces valid NAC sets. I'd send it to a stochastic-programming or combinatorial-optimization reviewer, not desk reject. A revised version that fixes the theorem would be a useful paper; the version now is a useful algorithm with an incorrect proof.

Referee Report

2 major / 5 minor

Summary. The paper proposes a graph-based algorithm (SNAC) for generating a set of non-anticipativity constraints (NACs) in multistage stochastic programs with gradual uncertainty realization and incomplete scenario sets. Scenarios are vertices and NACs are edges; Lemma 2 characterizes feasibility of a NAC set as connectivity of every block of every partition induced by each permissible event set. Algorithm 1 iterates from fine to coarse event sets, adding spanning-tree edges to connect each block, and Theorem 3 claims that the output is a minimum-cardinality edge set with running time O(|C||S|^3). The paper illustrates the algorithm on a manufacturing example and reports computational experiments on clinical-trial planning instances, comparing the SNAC-reduced formulation with the full pairwise NAC formulation.

Significance. The paper addresses a relevant and practically motivated problem: reducing the number of non-anticipativity constraints in multistage stochastic programs with gradual uncertainty realization and incomplete scenario sets. Its reformulation of the NAC-enforcement condition as block connectivity (Lemma 2) is a useful observation, and the SNAC algorithm is simple and clearly structured, with a plausible polynomial bound when the number of uncertain parameters is fixed. The computational case study demonstrates that the SNAC-based formulation can substantially reduce model generation and solution times relative to the full pairwise formulation, and the observation that SNAC can be run once for a family of instances sharing a scenario structure is a genuine practical advantage. However, the central theoretical result, the assertion of minimum-cardinality output, is not established and is in fact false for the algorithm as stated. If corrected and properly validated, the idea could serve as a useful heuristic; as it stands, the main contribution of the paper is not supported.

major comments (2)

[Section 3.1, Theorem 3 and Algorithm 1] The optimality claim in Theorem 3 is false as stated. Consider S={1,2,3,4,5,6}, C={c1,c2} with |c1|=|c2|=1, Π(c1)={A={1,5,6}, B={2,3,4}}, and Π(c2)={C={1,2,3}, D={4,5,6}}. This instance is realizable in the paper's framework (e.g., the manufacturing example of Section 2.2 with two products and the cuts corresponding to the first completion event for each product). A feasible set must connect A, B, C, and D. The minimum cardinality is 6: the edges (5,6),(1,5),(2,3),(2,4),(1,2),(4,5) achieve this, and no 5-edge set can, because the four 3-vertex blocks require at least 8 block-edge incidences while only the two edges (5,6) and (2,3) can be shared between two blocks, giving at most 7 incidences with 5 edges. If Algorithm 1 processes c1 first and at step 6 chooses the spanning trees (1,5),(1,6) on A and (2,4),(3,4) on B, then c2 has no previously added edges inside C or D, and the algorithm adds (1,2),(2,3),(4,5),(5,6), for a total of 8 edges. If instead c1's trees are (1,5),(5,6) and (2,3),(2,4), then c2 needs only (1,2) and (4,5), for a total of 6 edges. Since Algorithm 1 does not specify how the spanning tree at step 6 is constructed, both outputs are valid executions, and the theorem's 'minimum number of pairs' claim is contradicted. The proof's assertion that the spanning-tree choice 'has no effect on the number of arcs needed to be included at future iterations' is false: the choice at c1 changes the number of edges required at the same-iteration cut c2. More generally, the proof's necessity argument establishes only inclusion-minimality relative to the previously added edges, not minimum cardinality, and it does not account for alternate paths created by edges added later or at the same iteration.
[Section 4, Tables 2 and 3] The computational study does not test the optimality claim. Tables 2 and 3 compare the SNAC formulation only with the full pairwise NAC formulation, reporting reductions in the number of scenario pairs and in model generation and CPLEX times. No comparison is made against an exact minimum-cardinality NAC set (e.g., the MILP of Hooshmand and MirHassani (2016) or an exhaustive enumeration on the smallest instances), and no non-trivial lower bound is computed. The observed reductions are consistent with a heuristic that generates a feasible but not necessarily minimum set, so the experiments cannot substitute for a valid proof of Theorem 3.

minor comments (5)

[Section 3.1, Algorithm 1 and Remark 1] The treatment of cuts of the same cardinality is ambiguous: the for-loop in Algorithm 1 updates N after each c, so a later cut at the same level sees edges added for an earlier cut, whereas Remark 1 states that components are treated 'in parallel' without regard to edges added at the same iteration. This distinction matters for correctness, and the manuscript should specify the intended semantics and state the corresponding proof obligations.
[Section 3.2] The formula for the number of pairs in the full formulation is typeset with an extra space as '0 .5[|S|(|S|− 1)]'.
[Section 4] There are several typos, including 'suprizing' for 'surprising', 'teh' for 'the', and 'allied' for 'applied' in Remark 2.
[Abstract and Section 3.1] The abstract's statement that the algorithm 'scales polynomially in the number of scenarios' should carry the qualifier 'for a fixed number of uncertain parameters and events' that appears later in Section 3.1, since the running time is O(|C||S|^3) and |C| can be exponential in the number of events.
[Section 2.1] The displayed definition of the permissible event set C contains a malformed union expression; it should be written as a standard set comprehension.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: SNAC is a self-contained combinatorial construction whose optimality gap is a proof-validity concern, not an input-output loop.

full rationale

The paper's central claim is that Algorithm 1 generates a minimum-cardinality set of non-anticipativity constraints for multistage stochastic programs with incomplete scenario sets and gradual uncertainty realization. The derivation is a direct graph-theoretic construction: Lemma 2 characterizes sufficient NAC sets as those that make every scenario subset S^c_j connected in the induced subgraph, and Algorithm 1 greedily builds spanning trees over these subsets from the finest cuts downward. No parameter is fitted to data, no prediction is statistically forced, and no external result is imported from the authors' prior work to define the output. The cited prior work by Christian, Cremaschi, and Zeng appears only as background on decomposition heuristics and is not load-bearing for the optimality proof. The proof of Theorem 3 instead relies on Lemma 1, Lemma 2, and the construction of the algorithm itself. The only passage that warrants scrutiny is the unproven assertion in the proof of Theorem 3 that the choice of spanning tree for a cut 'has no effect on the number of arcs needed to be included at future iterations.' This is a potential correctness gap in the minimality argument, because equal-cardinality cuts need not be nested and the proof establishes inclusion-minimality rather than cardinality-minimality. However, that is a mathematical proof-validity problem, not circularity: the theorem's conclusion is not assumed in its premises, and the algorithm's output is not defined as 'the optimum' by construction. The computational study compares SNAC against the full pairwise NAC formulation, which tests constraint reduction and solution-time savings but not optimality; again, this is an empirical-validation limitation, not a circular derivation. No self-definitional, fitted-input, self-citation, imported-uniqueness, or ansatz-smuggling pattern is present. The algorithm is deterministic from the scenario set and event structure, and its claimed optimality is an independent combinatorial statement, even if that statement may require a more careful proof.

Assumptions & free parameters 0 free parameters · 2 assumptions · 0 invented entities

The algorithm has no fitted numerical parameters and introduces no new physical or mathematical entities. It relies on the domain assumption that event-set partitions accurately represent indistinguishability, and on the connectivity criterion for NACs. The least justified premise, the invariance of greedy spanning-tree choices across non-nested cuts, is treated as a red flag rather than an axiom.

assumptions (2)

domain assumption The information structure is fully captured by a finite set C of permissible event sets, and each event set c in C induces a partition of the scenario set into groups that are indistinguishable given the realized events.
Section 2.1 defines C and the partitions; the algorithm and its correctness rely on these partitions being known and complete for the MSSP.
domain assumption Non-anticipativity for a set of scenarios is equivalent to enforcing connectivity of the graph restricted to each cut subset, using transitivity of equality constraints.
Lemma 2 in Section 3.1 states this as a structural property of the MSSP formulation. It is asserted by construction rather than derived from an external theorem.

how reviews work

0 comments

Cite this review

Pith. "Pith review of A Graph Theoretic Approach to Non-Anticipativity Constraint Generation in Multistage Stochastic Programs with Incomplete Scenario Sets." pith.science (2026). https://pith.science/paper/IAECNQVY

@misc{pith2026190801792,
  author       = {Pith},
  title        = {Pith review of: A Graph Theoretic Approach to Non-Anticipativity Constraint Generation in Multistage Stochastic Programs with Incomplete Scenario Sets},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IAECNQVY}},
  note         = {Machine review of arXiv:1908.01792}
}

read the original abstract

We propose an algorithm for generating a minimum-cardinality set of non-anticipativity constraints (NAC) for scenario-based multistage-stochastic programming (MSSP) problems with both endogenous and exogenous uncertainties which allow for gradual realizations. Recently several authors have considered approaches to generate the minimum cardinality NAC set for MSSPs for various scenario set structures. However, these approaches have been limited to uncertain parameters where the realizations occur instantaneously or the full set of scenarios is required. The proposed algorithm, referred to as Sample Non-Anticipativity Constraint algorithm (SNAC) relaxes this requirement. We show that as long as the number of uncertain parameters and parameter values are kept constant, the algorithm scales polynomially in the number of scenarios.

Figures

Figures reproduced from arXiv: 1908.01792 by the authors.

**Figure 2.** The sets formed by each event set c ∈ C for the manufacturing example. 9 [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗

**Figure 3.** The sets formed by each event set c ∈ C for the manufacturing example with exogenous demand uncertainty. 12 [PITH_FULL_IMAGE:figures/full_fig_p012_3.png] view at source ↗

Figures from the paper (4 more)

**Figure 4.** Figure 4: The six sampled scenarios [PITH_FULL_IMAGE:figures/full_fig_p013_4.png]

**Figure 5.** Figure 5: Subsets formed by the event sets of length five [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]

**Figure 6.** Figure 6: Subsets for the iteration where k = 4 4. Computational Studies In this section we present results of a computational study. The aim of the study was to evaluate whether the proposed algorithm is able to significantly reduce the number of scenario pairs requiring enforc…

**Figure 7.** Figure 7: The last four iterations of the SNAC algorithm for the two product manufacturing example [PITH_FULL_IMAGE:figures/full_fig_p015_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 27 canonical work pages

[1]

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

ENTRY address author booktitle chapter edition editor howpublished institution 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 'mid.sentence := ...
[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 bbl.in capitalize ":" * " " *...
[3]

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 bbl.in ":" * " " * FUNCTION f...
[4]

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 bbl.in ":" * " " * FUNCTION f...
[5]

M., Grossmann, I

Apap, R. M., Grossmann, I. E., 2017. Models and computational strategies for multistage stochastic programming under endogenous and exogenous uncertainties. Computers & Chemical Engineering 103, 233--274

work page 2017
[6]

R., Louveaux, F., 2011

Birge, J. R., Louveaux, F., 2011. Introduction to stochastic programming. Springer Science & Business Media

work page 2011
[7]

A multistage stochastic programming approach to open pit mine production scheduling with uncertain geology

Boland, N., Dumitrescu, I., Froyland, G., 2008. A multistage stochastic programming approach to open pit mine production scheduling with uncertain geology. Optimization online, 1--33

work page 2008
[8]

Minimum cardinality non-anticipativity constraint sets for multistage stochastic programming

Boland, N., Dumitrescu, I., Froyland, G., Kalinowski, T., 2016. Minimum cardinality non-anticipativity constraint sets for multistage stochastic programming. Mathematical Programming 157 (1), 69--93

work page 2016

Show all 34 references

[9]

Heuristic solution approaches to the pharmaceutical r&d pipeline management problem

Christian, B., Cremaschi, S., 2015. Heuristic solution approaches to the pharmaceutical r&d pipeline management problem. Computers & Chemical Engineering 74, 34--47. ://www.sciencedirect.com/science/article/pii/S0098135414003433

2015
[10]

Variants to a knapsack decomposition heuristic for solving r&d pipeline management problems

Christian, B., Cremaschi, S., 2017. Variants to a knapsack decomposition heuristic for solving r&d pipeline management problems. Computers & Chemical Engineering 96, 18--32. ://www.sciencedirect.com/science/article/pii/S0098135416303271

2017
[11]

A branch and bound algorithm to solve large-scale multistage stochastic programs with endogenous uncertainty

Christian, B., Cremaschi, S., 2018 a . A branch and bound algorithm to solve large-scale multistage stochastic programs with endogenous uncertainty. AIChE Journal 64 (4), 1262--1271

2018
[12]

A multistage stochastic programming formulation to evaluate feedstock/process development for the chemical process industry

Christian, B., Cremaschi, S., 2018 b . A multistage stochastic programming formulation to evaluate feedstock/process development for the chemical process industry. Chemical Engineering Science 187, 223--244. ://www.sciencedirect.com/science/article/pii/S0009250918302148

2018
[13]

T., 2008

Colvin, M., Maravelias, C. T., 2008. A stochastic programming approach for clinical trial planning in new drug development. Computers & Chemical Engineering 32 (11), 2626--2642

2008
[14]

T., 2009

Colvin, M., Maravelias, C. T., 2009. Scheduling of testing tasks and resource planning in new product development using stochastic programming. Computers & Chemical Engineering 33 (5), 964--976. ://www.sciencedirect.com/science/article/B6TFT-4TK47C6-1/2/a31964976caa32338309a80...

2009
[15]

T., 2010 a

Colvin, M., Maravelias, C. T., 2010 a . Modeling methods and a branch and cut algorithm for pharmaceutical clinical trial planning using stochastic programming. European Journal of Operational Research 203 (1), 205--215. ://www.sciencedirect.com/science/article/B6VCT-4WXBM84-1...

2010
[16]

T., 2010 b

Colvin, M., Maravelias, C. T., 2010 b . Modeling methods and a branch and cut algorithm for pharmaceutical clinical trial planning using stochastic programming. European Journal of Operational Research 203 (1), 205--215

2010
[17]

A new class of planning models for option valuation of storage technologies under decision-dependent innovation uncertainty

Giannelos, S., Konstantelos, I., Strbac, G., 2017. A new class of planning models for option valuation of storage technologies under decision-dependent innovation uncertainty. In: 2017 IEEE Manchester PowerTech. pp. 1--6

2017
[18]

E., 2004

Goel, V., Grossmann, I. E., 2004. A stochastic programming approach to planning of offshore gas field developments under uncertainty in reserves. Computers & chemical engineering 28 (8), 1409--1429

2004
[19]

E., Sep 2006

Goel, V., Grossmann, I. E., Sep 2006. A class of stochastic programs with decision dependent uncertainty. Mathematical Programming 108 (2), 355--394. ://doi.org/10.1007/s10107-006-0715-7

2006 doi
[20]

E., El-Bakry, A

Goel, V., Grossmann, I. E., El-Bakry, A. S., Mulkay, E. L., 2006. A novel branch and bound algorithm for optimal development of gas fields under uncertainty in reserves. Computers & Chemical Engineering 30 (6-7), 1076--1092. ://www.sciencedirect.com/science/article/B6TFT-4JVTC...

2006
[21]

E., 2014

Gupta, V., Grossmann, I. E., 2014. A new decomposition algorithm for multistage stochastic programs with endogenous uncertainties. Computers & Chemical Engineering 62 (0), 62--79. ://www.sciencedirect.com/science/article/pii/S0098135413003621

2014
[22]

Efficient constraint reduction in multistage stochastic programming problems with endogenous uncertainty

Hooshmand, F., MirHassani, S., 2016. Efficient constraint reduction in multistage stochastic programming problems with endogenous uncertainty. Optimization Methods and Software 31 (2), 359--376. ://dx.doi.org/10.1080/10556788.2015.1088850

2016
[23]

A., 2016

Hooshmand Khaligh, F., MirHassani, S. A., 2016. A mathematical model for vehicle routing problem under endogenous uncertainty. International Journal of Production Research 54 (2), 579--590. ://doi.org/10.1080/00207543.2015.1057625

2016
[24]

W., Wets, R

Jonsbr ten, T. W., Wets, R. J.-B., Woodruff, D. L., Aug 1998. A class of stochastic programs withdecision dependent random elements. Annals of Operations Research 82 (0), 83--106. ://doi.org/10.1023/A:1018943626786

1998 doi
[25]

H., Woodruff, D

Jorjani, S., Scott, C. H., Woodruff, D. L., 1999. Selection of an optimal subset of sizes. International journal of production research 37 (16), 3697--3710

1999
[26]

Amsaa : A Multistep Anticipatory Algorithm for Online Stochastic Combinatorial Optimization

Mercier, L., Van Hentenryck, P., 2008. Amsaa : A Multistep Anticipatory Algorithm for Online Stochastic Combinatorial Optimization. pp. 173--187. ://dx.doi.org/10.1007/978-3-540-68155-7_15

2008 doi
[27]

Efficient solution procedures for multistage stochastic formulations of two problem classes

Solak, S., 2007. Efficient solution procedures for multistage stochastic formulations of two problem classes. Ph.D. thesis, GEORGIA INSTITUTE OF TECHNOLOGY

2007
[28]

B., Johnson, E

Solak, S., Clarke, J.-P. B., Johnson, E. L., Barnes, E. R., 2010. Optimization of r&d project portfolios under endogenous uncertainty. European Journal of Operational Research 207 (1), 420 -- 433. ://www.sciencedirect.com/science/article/pii/S0377221710003516

2010
[29]

E., 2008

Tarhan, B., Grossmann, I. E., 2008. A multistage stochastic programming approach with strategies for uncertainty reduction in the synthesis of process networks with uncertain yields. Computers & Chemical Engineering 32 (4-5), 766--788. ://www.sciencedirect.com/science/article/...

2008
[30]

E., Goel, V., 2009

Tarhan, B., Grossmann, I. E., Goel, V., 2009. Stochastic programming approach for the planning of offshore oil or gas field infrastructure under decision-dependent uncertainty. industrial & Engineering Chemistry Research 48 (6), 3078--3097. ://dx.doi.org/10.1021/ie8013549

2009 doi
[31]

E., Goel, V., 2013

Tarhan, B., Grossmann, I. E., Goel, V., 2013. Computational strategies for non-convex multistage minlp models with decision-dependent uncertainty and gradual uncertainty resolution. Annals of Operations Research 203 (1), 141--166. ://www.scopus.com/inward/record.url?eid=2-s2.0...

2013
[32]

Decision rules for information discovery in multi-stage stochastic programming

Vayanos, P., Kuhn, D., Rustem, B., 2011. Decision rules for information discovery in multi-stage stochastic programming. In: Proceedings of the IEEE Conference on Decision and Control. pp. 7368--7373. ://www.scopus.com/inward/record.url?eid=2-s2.0-84860672073&partnerID=40&md5=...

2011
[33]

A generalized knapsack-problem based decomposition heuristic for solving multistage stochastic programs with endogenous and or exogenous uncertainties

Zeng, Z., Christian, B., Cremaschi, S., 2018. A generalized knapsack-problem based decomposition heuristic for solving multistage stochastic programs with endogenous and or exogenous uncertainties. industrial & Engineering Chemistry Research 57 (28), 9185--9199. ://doi.org/10....

2018 doi
[34]

Artificial lift infrastructure planning for shale gas producing horizontal wells

Zeng, Z., Cremaschi, S., 2017. Artificial lift infrastructure planning for shale gas producing horizontal wells. Proceedings of the FOCAPO/CPC, Tuscan, AZ, USA, 8--12

2017

Pith tools

Reviewed August 14, 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 edition editor howpublished institution 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 'mid.sentence := ...

[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 bbl.in capitalize ":" * " " *...

[3] [3]

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 bbl.in ":" * " " * FUNCTION f...

[4] [4]

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 bbl.in ":" * " " * FUNCTION f...

[5] [5]

M., Grossmann, I

Apap, R. M., Grossmann, I. E., 2017. Models and computational strategies for multistage stochastic programming under endogenous and exogenous uncertainties. Computers & Chemical Engineering 103, 233--274

work page 2017

[6] [6]

R., Louveaux, F., 2011

Birge, J. R., Louveaux, F., 2011. Introduction to stochastic programming. Springer Science & Business Media

work page 2011

[7] [7]

A multistage stochastic programming approach to open pit mine production scheduling with uncertain geology

Boland, N., Dumitrescu, I., Froyland, G., 2008. A multistage stochastic programming approach to open pit mine production scheduling with uncertain geology. Optimization online, 1--33

work page 2008

[8] [8]

Minimum cardinality non-anticipativity constraint sets for multistage stochastic programming

Boland, N., Dumitrescu, I., Froyland, G., Kalinowski, T., 2016. Minimum cardinality non-anticipativity constraint sets for multistage stochastic programming. Mathematical Programming 157 (1), 69--93

work page 2016