Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Collections

The OR Formalization Drive

A coordinated effort to formalize the mathematics of operations research in Lean 4.

Scope: missions in linear and integer programming, convex and nonlinear optimization, combinatorial optimization, dynamic programming, scheduling, stochastic models such as queueing and inventory, simulation, revenue management, and game theory and mechanism design.

1285 open missions

Missions

451–475 of 1285
Open1285Completed3527All4812
Convex OptimizationMachine LearningOptimization·Captain: mikedeng1

Introduction to Online Convex Optimization VI: Bandit Convex Optimization via Gradient EstimationTextbook

Every algorithm in Chapters I–V observes the full cost function ftf_tft​ after playing xtx_txt​. Many applications only reveal the scalar cost ft(xt)f_t(x_t)ft​(xt​) incurred — routing a network and observing total latency, or placing an ad and observing the click-through revenue, without ever seeing the cost of a path or bid not taken. This is the bandit feedback model, and…

10 thms3 active users
Machine LearningReinforcement LearningStatistics·Captain: mikedeng1

Foundations of Reinforcement Learning V: General Decision Making and the Decision-Estimation Coefficient Lower BoundTextbook

Online decision-making problems — multi-armed bandits, contextual bandits, structured bandits, and episodic reinforcement learning — look superficially different but share a common shape: a learner repeatedly acts, observes feedback, and is scored by regret against the best action in hindsight. Foster, Kakade, Qian and Rakhlin's Foundations of Reinforcement…

20 thms3 active users
Control TheoryDynamic ProgrammingProbability+1·Captain: iris

Markov Decision Processes with Applications to Finance XIII: Piecewise Deterministic Markov Decision Processes Have Optimal Nonrelaxed Markov PoliciesTextbook

Many controlled systems move deterministically between random shocks: an insurer's reserve between claims, a portfolio between market jumps, a production line between breakdowns. Piecewise Deterministic Markov Processes, introduced by Davis (Piecewise-deterministic Markov processes, JRSS B, 1984), are the general class of such non-diffusions. Their control…

21 thms2 active usersReviewed
Operations ResearchOptimizationProbability·Captain: iris

Good Prophets Know When the End Is Near: A Heads-Up T^(1/2+ε) Periods Before the End Gives Expected Loss Against the Clairvoyant Optimum Bounded Independently of the HorizonResearch Paper

Many online decision problems in operations research have the same shape. Requests of finitely many types arrive one per period, and each must be answered at once by an irrevocable action: an item is packed into a bin configuration, a payment is routed to a processor, a booking is accepted or rejected. The cost of the whole run depends only on how many…

14 thms2 active usersReviewed
EconomicsOperations Research·Captain: iris

Product Quality and Information Sharing in the Presence of Reviews: perfectly informative reviews weaken the quality incentive under a wholesale price contract but not under a commission contractResearch Paper

Online marketplaces hold two kinds of information that a manufacturer selling through them would like to have. Product reviews tell customers about a product's quality. Customer data tell the platform how strongly the market responds to quality, and the platform can sell that signal to the manufacturer, as Amazon does with Brand Analytics, which Shin and…

18 thms2 active usersReviewed
Dynamic ProgrammingLinear OptimizationOperations Research·Captain: iris

Self-Adapting Network Relaxations for Weakly Coupled Markov Decision Processes: Every Feasibility Network Relaxation Bound Lies Between the Optimal Value and the Lagrangian BoundResearch Paper

Many sequential decision problems in operations research consist of a number of independent subsystems that interact only through shared resources: assortment and inventory decisions under a common budget, multi-armed bandits with a cap on the number of arms pulled, scheduling with capacity limits. Such problems are modelled as weakly coupled Markov…

14 thms2 active usersReviewed
Operations ResearchProbabilityRevenue Management and Pricing·Captain: iris

Technical Note—A Note on State-Independent Policies in Network Revenue Management: the make-to-stock policy has optimality gap O(k)O(\sqrt{k})O(k​) under non-separable demandResearch Paper

An airline, a hotel chain or an online retailer sells many products that draw on a shared, limited stock of resources: seats on flight legs, rooms on nights, units of a component. Choosing prices in real time from the current stock is the network revenue management problem, and it is intractable in general. A practical answer is a state-independent policy,…

15 thms2 active usersReviewed
Linear OptimizationOperations ResearchOptimization+1·Captain: iris

Degeneracy Is OK: Logarithmic Regret for Network Revenue Management with Indiscrete Distributions: Algorithms 1+2 Have Regret ≤((2κ1+2)/α+(4/α)∑j1/pj)log⁡2T+s0rmax⁡\le((2\kappa_1+2)/\alpha+(4/\alpha)\sum_j1/p_j)\log^2T+s_0r_{\max}≤((2κ1​+2)/α+(4/α)∑j​1/pj​)log2T+s0​rmax​Research Paper

Network revenue management (NRM) is the problem of selling several limited resources to customers who arrive one at a time and must be accepted or rejected on the spot. Examples are airline seats over flight legs, hotel rooms over nights, and cloud capacity over servers. Each request uses a fixed bundle of resources and offers a reward. The seller wants to…

25 thms2 active usersReviewed
Dynamic ProgrammingLinear OptimizationOperations Research+1·Captain: iris

Technical Note—Revenue Management with Calendar-Aware and Dependent Demands: Asymptotically Tight Fluid Approximations: The Fluid-LP Policy Is 1/(4L)-Approximate and Asymptotically OptimalResearch Paper

Revenue management systems forecast demand over canonical intervals of time, such as a week, often with a general distribution for each week, and the demand of one week is a signal for the next. The prevalent demand model, a discrete-time approximation to a Poisson process, cannot reproduce this. With one arrival per period, arrival probabilities λt\lambda_tλt​ a…

22 thms2 active usersReviewed
Linear OptimizationOperations ResearchOptimization+1·Captain: iris

Effective Online Order Acceptance Policies for Omni-Channel Fulfillment: The Expected Total Cost Is Unimodal in the Global ThresholdResearch Paper

In omni-channel retail, a chain's stores double as shipping centres for online orders. Online orders arrive over a period and must be accepted or rejected as they come in. They are filled at the end of the period from whatever store inventory walk-in customers have left. An accepted order that cannot be filled is cancelled at a cost. A rejected order that…

25 thms2 active usersReviewed
Algorithmic Game TheoryMachine LearningMechanism Design+1·Captain: iris

Liquid Welfare Guarantees for No-Regret Learning in Sequential Budgeted Auctions: γ-Competitive Buyers in First-Price Auctions Get Liquid Welfare ≥ (LW* − O(n)(Reg+1))/(γ + 1/2 + O(1/γ))Research Paper

Most online advertising is sold through sequential auctions: one impression at a time, millions of times a day, and since 2019 increasingly by first-price rules (Won21). The advertisers who bid are budget-constrained, and they do not compute equilibria; their bids are set by learning algorithms that adjust a shading or pacing multiplier from round to round.…

27 thms2 active usersReviewed
CombinatoricsOperations ResearchOptimization+2·Captain: iris

Online Bipartite Matching with Reusable Resources: Periodic Reranking with β=0.89\beta = 0.89β=0.89 Is 0.58930.58930.5893-Competitive for Vertex-Weighted Reusable MatchingResearch Paper

Many allocation platforms match each arriving request to a compatible resource that is used for a while and then returns: virtual machines in cloud computing, listings on rental marketplaces, hospital beds and ventilators. The online bipartite matching problem of Karp, Vazirani and Vazirani (1990) models the case where each resource can be used only once.…

31 thms2 active usersReviewed
Algorithmic Game TheoryCombinatoricsOperations Research+2·Captain: iris

The Secretary Problem with Independent Sampling: In the Adversarial-Order Problem with p-Sampling, the k-Max Algorithm with k = ⌊1/(1−p)⌋ Has the Best Possible Success Guarantee kpᵏ(1−p)Research Paper

The classical secretary problem asks a player to stop on the best of nnn items that arrive one at a time; with a uniformly random order the optimal success probability is 1/e1/e1/e, and with an adversarial order nothing nontrivial is possible. Much recent work in online algorithms and mechanism design asks how much prior information restores good guarantees.…

21 thms2 active usersReviewed
Dynamic ProgrammingEconomicsOperations Research·Captain: iris

Pricing for the Stars: Dynamic Pricing in the Presence of Rating Systems: rating sensitivity raises long-run profits and ratings, and raises prices and lowers consumer surplus exactly when κ < 1/2Research Paper

Online marketplaces such as Amazon and Steam show each product's aggregate rating, a summary of past reviews, next to its current price. Reviews are left by buyers, and who buys depends on the price, so a seller who changes the price today changes the reviews that tomorrow's rating is made of. Platforms also choose how the rating is computed; in recent…

8 thms2 active usersReviewed
Operations ResearchOptimizationProbability·Captain: iris

Overbooking with Bounded Loss: The Online Index Policy's Expected Loss Against the Clairvoyant Optimum Is Bounded by a Constant Independent of Capacity and HorizonResearch Paper

Airlines, hotels and clinics sell more reservations than they have capacity because some customers do not show up. Accepting too few leaves capacity idle; accepting too many forces the firm to compensate customers it cannot serve. When customers differ both in the revenue they bring and in their probability of showing up, the firm must decide online, one…

25 thms2 active usersReviewed
Differential GeometryNumerical AnalysisOptimization·Captain: iris

A Riemannian Smoothing Steepest Descent Method for Non-Lipschitz Optimization on Submanifolds: RSSD Accumulation Points Are Stationary, and Limiting Stationary under Gradient Sub-ConsistencyResearch Paper

Many estimation problems in signal processing and statistics ask for a sparse or robust solution under an orthogonality or normalization constraint: finding the sparsest vector in a linear subspace over the unit sphere, or learning an orthogonal dictionary over the orthogonal group. Sparsity is promoted by penalties such as ∥x∥pp\|x\|_p^p∥x∥pp​ with 0<p<10<p<10<p<1, which are no…

15 thms2 active usersReviewed
Graph TheoryOperations ResearchProbability·Captain: iris

On (Random-order) Online Contention Resolution Schemes for the Matching Polytope of (Bipartite) Graphs: a 0.474-Selectable Random-Order Contention Resolution Scheme for Every GraphResearch Paper

Contention resolution is a standard way to turn a fractional solution into an integral one while losing only a constant factor on every coordinate. For matchings on graphs it underlies guarantees for stochastic matching, prophet inequalities, and posted-price mechanisms (Feldman, Svensson and Zenklusen 2021; Ezra, Feldman, Gravin and Tang 2022). A scheme…

24 thms2 active usersReviewed
Numerical AnalysisOperations ResearchOptimization·Captain: iris

Worst-Case Iteration Bounds for Log Barrier Methods on Problems with Nonconvex Constraints: Trust-IPM Reaches an Approximate Second-Order SIP in O(1 + (ψµ(x⁰) − ψ*µ)/µ · (L₁/(µτₗ²))^{3/4}) IterationsResearch Paper

Interior point methods (IPMs) are the basis of widely used nonlinear optimization solvers, and in practice they are applied to problems whose constraints are nonconvex. For linear and conic programs the theory is complete: self-concordance gives an O(clog⁡(1/ϵ))O(\sqrt{c}\log(1/\epsilon))O(c​log(1/ϵ)) iteration bound. For nonconvex constraints, the analyses of practical IPMs show…

16 thms2 active usersReviewed
Mechanism DesignOperations ResearchProbability·Captain: iris

Distributionally Robust Optimal Allocation with Costly Verification: Under Markov Ambiguity, a Type-(ii) Favored-Agent Mechanism with Threshold at Its Maximal Type Is Pareto Robustly OptimalResearch Paper

A principal has one indivisible good to give to one of several agents. Each agent would put the good to a use whose value to the principal is that agent's private type; money cannot change hands, but the principal can inspect an agent at a cost and so learn his true type. Allocation problems of this kind arise wherever a budget, a slot or a license is…

24 thms2 active usersReviewed
Convex OptimizationOperations ResearchOptimization·Captain: iris

On Degenerate Doubly Nonnegative Projection Problems: The Dual Problem Has Quadratic Growth at Every Dual Solution When rank X̄ ≥ n − 1 or a Strictly Complementary Dual Solution ExistsResearch Paper

The doubly nonnegative (DNN) cone Dn={X∈Sn∣X⪰0, X≥0 entrywise}\mathbb D^n=\{X\in\mathbb S^n\mid X\succeq0,\ X\ge0\text{ entrywise}\}Dn={X∈Sn∣X⪰0, X≥0 entrywise} is the standard tractable relaxation of the completely positive cone, which appears in copositive reformulations of quadratic and combinatorial optimization problems. Projecting a symmetric matrix onto Dn\mathbb D^nDn is a basic subroutine of first-order…

16 thms2 active usersReviewed
Discrete GeometryLinear OptimizationOperations Research+1·Captain: iris

Proximity and Flatness Bounds for Linear Integer Optimization: With Integral bbb, an Optimal LP Vertex Is Strictly Within ℓ∞\ell_\inftyℓ∞​-Distance 4n+29Δn−1(A)\frac{4n+2}{9}\Delta_{n-1}(A)94n+2​Δn−1​(A) of an Integer OptimumResearch Paper

An integer linear program asks for the maximum of c⊤xc^\top xc⊤x over the integer points of a polyhedron P(A,b)={x∈Rn:Ax≤b}P(A,b)=\{x\in\mathbb R^n: Ax\le b\}P(A,b)={x∈Rn:Ax≤b}. Its linear relaxation is much easier to solve, so a basic question is how far an optimal vertex x∗x^*x∗ of the relaxation can be from the nearest optimal integer solution z∗z^*z∗. A bound on ∥x∗−z∗∥∞\|x^*-z^*\|_\infty∥x∗−z∗∥∞​ confines the search…

22 thms2 active usersReviewed
EconomicsOperations ResearchOptimization·Captain: iris

Dynamic Pricing with Online Reviews: Under Quality-Based Reviews, a Fixed Price Is Asymptotically Optimal Except at the Critical Time Scale γT≍1\gamma T \asymp 1γT≍1Research Paper

When a new product goes on sale, customers do not know its quality, and each purchase can produce a review that changes what later customers believe. A seller's price therefore does two things: it earns revenue now, and it sets the pace at which information reaches the market. Whether this second effect justifies changing the price over time is a practical…

35 thms2 active usersReviewed
Operations ResearchOptimizationProbability·Captain: iris

ρ-Arbitrage and ρ-Consistent Pricing for Star-Shaped Risk Measures: Under the Fatou Property and Sensitivity to Large Expected Losses, No ρ-Arbitrage ⇔ No ρ∞-Arbitrage ⇔ ρ∞₁ > 0Research Paper

A risk measure condenses the risk of a financial position into one number, and regulators use such numbers to constrain what a financial agent may hold. Portfolio selection under a risk constraint goes back to the mean-variance problem of Markowitz (1952). Modern regulation works instead with monotone risk measures such as Value at Risk and Expected…

18 thms2 active usersReviewed
Algorithmic Game TheoryCombinatoricsMechanism Design+2·Captain: iris

The Core of Housing Markets from an Agent’s Perspective: Is It Worth Sprucing up Your Home?: The Core Respects Improvement, and an Allocation Proving It Is Found in Linear TimeResearch Paper

In a housing market, introduced by Shapley and Scarf (J. Math. Econ. 1974), every agent owns one indivisible good and the agents reallocate the goods among themselves. The model underlies kidney exchange programmes, where a patient–donor pair "owns" its donor and pairs exchange donors along cycles (Roth, Sönmez, Ünver, QJE 2004). A natural design…

23 thms2 active usersReviewed
Bandit AlgorithmsOperations ResearchStochastic Systems·Captain: iris

Efficient Decentralized Multi-agent Learning in Asymmetric Bipartite Queueing Systems: Under DAM.FE the Time-Averaged Weighted Queue Length Is O(K²(log N + K)L_check/ε³ + K·min(T,T₀)²/(εT))Research Paper

Many service systems match a set of agents with work to a set of servers in every time slot: devices sharing radio channels in cognitive radio, workers and tasks on an online service platform, customers and providers in a marketplace. When the matching is computed centrally and the service rates are known, the MaxWeight policy, which in every slot serves a…

22 thms2 active usersReviewed
PreviousPage 19 of 52Next

Top contributors

RankContributorAccepted solutionsSubmitted problems
1TItianyipeng5162497
2EYEyal199050327667
3MRmrfancypants41060
4WIWillR36415412
5NINickrobbins953286155
6VEvebis25800
7FIfizzyinc321725
8SHshivm17490
9SCShuze Chen9455324
10KMkmehltretter82895261

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me