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.
The Statistical Complexity of Interactive Decision Making IV: Log-Loss Regret Controls Hellinger Estimation Error, Est_H ≤ Reg_KL + 2 log(1/δ) with Probability 1−δResearch Paper
Interactive decision methods often need an estimate of an unknown outcome law after each round. A learner can be good at decisions only if its model of the observations is accurate where future decisions will sample. Foster, Kakade, Qian and Rakhlin measure this accuracy by a cumulative squared Hellinger distance and use it in their regret bounds for…
The Statistical Complexity of Interactive Decision Making IX: Posterior Sampling Bounds the Square-Loss DEC by 2Δ + 24·θ·log²(γ)/γ, θ the Disagreement CoefficientResearch Paper
In a structured bandit problem a learner repeatedly chooses a decision π from a set Π and observes a noisy reward whose mean fM(π) is an unknown function from a known class FM. Classical analyses of such problems measure the complexity of FM by the eluder dimension (Russo and Van Roy, 2013; 2014) or the st…
Finding Cliques in Social Networks: A New Distribution-Free Model 3: For Every c ≥ 2 and n ≥ c There Is a c-Closed Graph on n Vertices with Ω(c^(−3/2)·2^(c/2)·n^(3/2)) Maximal CliquesResearch Paper
Social networks exhibit triadic closure: two people with many common friends tend to be friends themselves. Fox, Roughgarden, Seshadhri, Wei and Wein (arXiv:1804.07431, SIAM J. Comput. 2020) turn this empirical regularity into a deterministic graph class, the c-closed graphs, and ask how hard a classical NP-hard problem, enumerating maximal cliques, becomes…
The Statistical Complexity of Interactive Decision Making VIII: Linear Bandits on the Unit Ball Have Localized DEC at Least d/(12γ)Research Paper
In a linear bandit a learner repeatedly chooses a decision π from a set Π⊆Rd and observes a noisy reward whose mean is ⟨θ⋆,π⟩ for an unknown parameter θ⋆. It is the standard model for sequential decisions with many correlated actions, such as ad placement, routing and recommendation, and its…
Potential Games Are Necessary to Ensure Pure Nash Equilibria in Cost Sharing Games 1: Every Game Has a Pure Nash Equilibrium Iff the Distribution Rules Are Generalized Weighted Shapley ValuesResearch Paper
In a cost sharing game (or, symmetrically, a revenue sharing game) self-interested players each pick a set of resources, every resource generates a welfare that depends on who uses it, and a distribution rule splits that welfare among the users. Network formation, atomic routing, coverage and distributed-control problems all fit this model (Anshelevich et…
Approximate Revenue Maximization with Multiple Items 1: For Independent Groups of Goods Y and Z, Rev(Y, Z) ≤ Rev(Y) + Rev(Z) + BRev(Y) + BRev(Z)Research Paper
A seller who offers several goods to a buyer with private values has to choose a selling mechanism. For one good the optimal mechanism is known: post the best take-it-or-leave-it price (Myerson 1981). For two or more goods the optimal mechanism can be complicated: it may use lotteries, its menu may be infinite, and it may not be computable in closed form…
Online Contention Resolution Schemes with Applications to Bayesian Selection Problems 2: The Degree Relaxation of Matching Has a (b, e^(−2b))-Selectable Randomized Greedy OCRSResearch Paper
In online selection, candidate items reveal themselves one at a time and an algorithm must decide immediately whether to keep each one. When the kept items must form a matching, accepting one edge can prevent a later edge sharing an endpoint from being accepted. A fractional solution supplies useful marginal probabilities, but it does not by itself specify…
Online Contention Resolution Schemes with Applications to Bayesian Selection Problems 5: A (b, c)-Selectable Greedy OCRS Rounds x ∈ bP to E[f(S)] ≥ c·F(x) for Monotone Submodular fResearch Paper
Many online selection problems ask an algorithm to accept or reject items one at a time, irrevocably, while keeping the accepted set feasible: independent in a matroid, a matching in a graph, or within a knapsack budget. In Bayesian versions, such as prophet inequalities and posted-price mechanisms, a fractional point x of a relaxation of the feasible sets…
Online Contention Resolution Schemes with Applications to Bayesian Selection Problems 3: A Knapsack Polytope Has a (b, (1−2b)/(2−2b))-Selectable Randomized Greedy OCRS for b ≤ 1/2Research Paper
In online selection, useful items arrive one at a time and must be accepted or rejected before the next arrival. A knapsack constraint limits the total size of accepted items to one unit. When the order may depend on all of the random outcomes, an algorithm needs a guarantee that survives every such order. Feldman, Svensson, and Zenklusen introduced online…
Generalized Turán Problems for Even Cycles 6: The Maximum Number of (2l+1)-Cycles in a Graph with No Cycle of Length 3, …, 2l or 2k+1 Is Ω(n^{1+1/(2k+2)}) and O(n²)Research Paper
Generalized Turán problems ask for the maximum number of copies of a fixed graph H in an n-vertex graph that avoids a family F of forbidden subgraphs. The ordinary Turán number counts edges (H=K2); Alon and Shikhelman (arXiv:1506.03397) started the systematic study of the general function ex(n,H,F), building on earlier…
Online Contention Resolution Schemes with Applications to Bayesian Selection Problems 1: Every Loopless Matroid Polytope Has a (b, 1−b)-Selectable Deterministic Greedy OCRSResearch Paper
In online selection, items reveal information one at a time, and each decision to accept an item is irrevocable. A matroid constraint permits some subsets and excludes others; it covers choices such as selecting linearly independent vectors or acyclic edges. A fractional solution can assign an acceptance probability to each item, but independent activations…
Exact Solution of the Hypergraph Turán Problem for k-Uniform Linear Paths: For k ≥ 4 and Large n, the Turán Numbers of P_{2t+1} and P_{2t+2} and Their Unique Extremal FamiliesResearch Paper
Hypergraph Turán problems ask how many edges a k-uniform hypergraph on n vertices can have without containing a fixed sub-hypergraph. For graphs (k=2) the theory is well developed; for k≥3 exact answers are rare and are known mostly for small, dense configurations such as the Fano plane or expanded triangles. Paths are the natural sparse test case. For…
Minimum Cost Flows in Graphs with Unit Capacities: When the Second Phase of a Cost Scale Starts, at Most 3λ Excess Is Left for λ = √m and 36λ for λ = n^(2/3)Research Paper
Minimum-cost flow is a standard model for routing a quantity through a network while paying a cost on each arc. Unit capacities arise when each input arc can carry at most one unit; they include flow formulations of several matching and assignment problems. Goldberg, Kaplan, Hed and Tarjan analyze a cost-scaling algorithm for this setting. Their…
Certifying Some Distributional Robustness with Principled Adversarial Training 2: The Empirical Robustness Level E_P̂ₙ[c(T_γ(θ;Z),Z)] Concentrates Uniformly over ΘResearch Paper
Neural networks trained by empirical risk minimization are sensitive to small adversarial perturbations of their inputs. Distributionally robust optimization addresses this by training against the worst distribution in a neighbourhood of the data distribution. Sinha, Namkoong, Volpi and Duchi (arXiv:1710.10571v5, ICLR 2018) take the neighbourhood to be a…
Input-to-State Stability of Nonlinear Impulsive Systems 2: An Exponential ISS-Lyapunov Function Makes an Impulsive System Uniformly ISS under the Generalized Average Dwell-Time ConditionResearch Paper
An impulsive system evolves continuously most of the time and changes state instantaneously at specified impulse times. Reset controllers, sampled interventions, and hybrid models have this shape. A stability estimate for the continuous dynamics alone can fail after many jumps: a jump may enlarge the state, and even a contracting jump can be too sparse to…
Certifying Some Distributional Robustness with Principled Adversarial Training 1: SGD on the Wasserstein Penalty Surrogate Finds an Approximate Stationary Point at Rate 4σ√(L_φΔ_F/T)Research Paper
Neural networks trained by empirical risk minimization are sensitive to small, deliberately chosen perturbations of their inputs ("adversarial examples"). A standard defence is adversarial training: at each step, perturb the training example to increase the loss, then take a gradient step at the perturbed point. Sinha, Namkoong, Volpi and Duchi (arXiv:1710.1…
Input-to-State Stability of Nonlinear Impulsive Systems 1: An ISS-Lyapunov Function Makes an Impulsive System ISS for Every Impulse Sequence Satisfying the Nonlinear Fixed Dwell-Time ConditionResearch Paper
Impulsive systems evolve continuously most of the time and change their state abruptly at isolated instants. They model sampled-data and networked control loops, where a controller acts at transmission times; population and epidemic models with periodic harvesting or vaccination; and mechanical systems with impacts. Their inputs are disturbances,…
Input-to-State Stability of Nonlinear Impulsive Systems 3: Small-Gain Theorem: max_i σ_i⁻¹(V_i(x_i)) Is an ISS-Lyapunov Function for an Interconnection of Impulsive SystemsResearch Paper
Large control systems are rarely analysed as a whole. A power grid, a logistics network or a chemical plant is a network of subsystems, each with its own model, coupled through their states. Input-to-state stability (ISS) is the standard notion of robustness for such systems: the state is bounded by a decaying function of the initial state plus a gain of…
Regret in Online Combinatorial Optimization II: Under Bandit Feedback Every Strategy Has Minimax Regret at Least 0.02 m√(dn) on Some Set of Actions with m OnesResearch Paper
An online player repeatedly chooses a subset of coordinates and pays the sum of their losses. In many applications the player sees only that sum, not the losses of the selected coordinates. This bandit feedback makes it harder to identify which coordinate within each chosen subset is responsible for a favorable or unfavorable outcome. Audibert, Bubeck and…
Recovering Best Statistical Guarantees via the Empirical Divergence-Based Distributionally Robust Optimization 3: On a Finite Support the Empirical DRO Equals the Burg-Divergence DRO with χ²₁ RadiusResearch Paper
Distributionally robust optimization (DRO) evaluates a decision against a collection of probability distributions near an empirical one. In a finite scenario model, the familiar construction puts a divergence ball around the histogram of observed scenarios. Lam's empirical construction instead gives one weight to each observation, even when several…
Prophet Inequalities Made Easy: Stochastic Optimization by Pricing Nonstochastic Inputs II: Posted Expected Weakly (α, β₁, β₂)-Balanced Prices Earn 1/(α(2β₁ + 4β₂)) of the Expected Welfare of ALGResearch Paper
A prophet inequality compares an online decision maker, who sees random inputs one at a time and must commit irrevocably, with a "prophet" who sees all inputs in advance. The classic single-item version (Krengel–Sucheston, Samuel-Cahn) says that a well-chosen fixed threshold earns at least half of the expected maximum. In mechanism design the same statement…
Data-Based Dynamic Pricing and Inventory Control with Censored Demand and Limited Price Changes 2: Any Policy with at Most m Price Changes Has Regret at Least KT^(1/(m+1)) on a Bernoulli InstanceResearch Paper
A seller learning a demand curve normally adjusts prices as new observations arrive. A price-change limit prevents that continual adjustment: a price may have to be chosen before the seller knows which demand curve it faces, then held through a block of sales. This matters when repricing is operationally costly, constrained by contracts, or visible to…
Optimization-Based Scenario Reduction for Data-Driven Two-Stage Stochastic Optimization II: Under a Symmetric Uncertainty Law the Mean Solves the Convex Upper-Bound Scenario ProblemResearch Paper
A two-stage stochastic programme chooses a decision z before an uncertain parameter ξ is revealed and pays a cost c(z;ξ). When ξ is known only through n data points, or through a law with many atoms, the expected cost is expensive to optimize, and practitioners replace the law by a small set of mreduced scenarios. Classical scenario reduction chooses…
Robust Dynamic Pricing with Strategic Customers: Over Horizon T the Simple Robust Pricing Policy Earns at Least V*_β(x₀)/(1 + e^{−βT}/(βT))Research Paper
A seller with a fixed stock must choose prices before knowing when customers will arrive or what they will be willing to pay. If customers can wait for a later price, the seller must also account for how today's price changes their purchasing decisions. Chen and Farias study this problem with strategic customers and identify a simple inventory-based policy…
Stochastic First- and Zeroth-Order Methods for Nonconvex Stochastic Programming 1: The Randomized Stochastic Gradient Method Has (1/L)E‖∇f(x_R)‖² ≤ (D_f² + σ²Σγ_k²)/Σ(2γ_k − Lγ_k²)Research Paper
Stochastic gradient methods are the standard tool for minimizing an objective that can only be accessed through noisy gradient samples. Training a machine-learning model on sampled data, simulation-based optimization, and stochastic programs with endogenous uncertainty are all of this kind. In many of these applications the objective is nonconvex: the loss…