The paper develops algorithms to minimize queue length regret in a communication system.
problem Minimizing the difference between actual and optimal queue lengths over time slots.
method Introduces queue length regret and applies algorithms from stochastic multi-armed bandit problem to analyze system performance.
result Order optimal O ( 1 ) O(1) O ( 1 ) queue length regret can be achieved with queue-length based policies. Study on queues with Hawkes arrivals, proving steady-state behavior and developing an efficient algorithm.
problem Analyzing the steady-state behavior of queues with Hawkes arrivals.
method Novel coupling techniques and exponential convergence results for workload and busy period processes.
result Exponential convergence of queueing processes to their stationary distribution.
Algorithm stabilizes queues in asymmetric systems with unknown service rates.
problem Stabilizing queues in multi-class multi-server systems with unknown service rates.
method Proposes UCB and Thompson Sampling algorithms to stabilize queues while learning service rates.
result Achieves system stability with an average queue length bound of \(O(\min\{N,K\}/ε)\) for large time horizon \(T\).
The paper analyzes transaction fees on blockchains using a priority queue model.
problem Understanding and optimizing transaction fees on blockchain networks.
method An M/G^K/1 priority queue model is used to analyze transaction fees and user behavior.
result New insights into the dynamics of transaction fees and their impact on user behavior are provided.
Designs for allocating resources to prioritize needy applicants while estimating treatment effects.
problem Resource allocation under uncertainty with prioritized queues.
method Priority-queue randomization for treatment assignment and estimation of treatment effects.
result Identification of causal effects under different arrival and treatment assignment scenarios.
Improved queue-reactive model considers order sizes for better market simulation.
problem Accurately modeling market dynamics and order flow properties.
method Integrates order sizes, type, and arrival rate into queue-reactive model.
result Extended model produces markets with volatility matching historical data.
MERLIN tackles multi-objective task scheduling with hierarchical DRL, outperforming existing methods.
problem Optimizing multiple conflicting constraints in multi-objective task scheduling with varying queue sizes.
method Hierarchical deep reinforcement learning approach to manage large queues efficiently.
result MERLIN outperforms existing methods by a large margin (>22%) on multiple queue sizes.
Enhances queue-reactive model for realistic limit order book simulation.
problem Realistic simulation of limit order books for market research and strategy development.
method Extends Queue-Reactive model with neural network for complex dependencies and varying market conditions.
result Captures key market properties like square-root law of market impact and order size patterns.
Modern Internet services, such as those at Google, Yahoo!, and Amazon, handle billions of requests per day on clusters of thousands of computers. Because these services operate under strict performance requirements, a statistical understanding of their performance is of great practical interest. Such services are model…
Two Hawkes models integrate queue sizes to model order flow and improve accuracy.
problem Modeling the stochastic time evolution of a limit order book.
method Queue-reactive Hawkes models with explicit queue size dependencies.
result Hawkes term significantly improves order flow description and queue distributions.
The paper models blockchain queues and trading dynamics, finding conditions for transaction priority and price impact.
problem Understanding and predicting price impacts in blockchain trading environments.
method Developed a probabilistic model for blockchain queues with adversarial scheduling, derived expressions for transaction priority and price impact.
result Conditions for transaction priority and statistical models for price impact in blockchain trading environments.
Queue-based resampling tackles online class imbalance learning with selective resampling of past examples.
problem Online class imbalance learning under class imbalance and concept drift.
method Queue-based resampling algorithm that selectively includes past examples in the training set.
result Queue-based resampling outperforms state-of-the-art methods in terms of learning speed and quality.
Order positions are key variables in algorithmic trading. This paper studies the limiting behavior of order positions and related queues in a limit order book. In addition to the fluid and diffusion limits for the processes, fluctuations of order positions and related queues around their fluid limits are analyzed. As a…
Study improves queue length estimation from connected vehicles by filtering parameters.
problem Large errors in estimated queue lengths at low market penetration rates.
method Used Kalman and Particle filters as multilevel real-time estimators.
result Filters reduce estimation errors and improve accuracy within 15 minutes.
Queue imbalance predicts mid-price movement direction.
problem Predicting mid-price movement direction in limit order books.
method Logistic regression and local logistic regression fits between queue imbalance and mid-price movement direction.
result Queue imbalance provides significant predictive power for mid-price movement direction.
Finite-time queue peaks in stochastic networks have logarithmic scaling after geometric thresholds.
problem Queue peak laws in stochastic networks with geometric thresholds.
method Self-normalization mechanism
result Logarithmic scaling of queue peaks after geometric thresholds.
The paper tackles ride-hailing fleet repositioning with a calibrated demand approach.
problem Repositioning idle supply before future demand is observed in ride-hailing.
method A predict-then-optimize approach using calibrated demand regimes, a similarity gate, and spatial queue-regret decomposition.
result The spatial gate reduces mean wait time to 82.3s compared to 85.3s for a hand-tuned similarity gate and 85.8s for a distributional-only baseline.
Motivated by empirical data, we develop a statistical description of the queue dynamics for large tick assets based on a two-dimensional Fokker-Planck (diffusion) equation, that explicitly includes state dependence, i.e. the fact that the drift and diffusion depends on the volume present on both sides of the spread. "J…
Paper uses queue theory to model financial signals with relativistic delay.
problem Relativistic delay in financial trading signals.
method Modified M/M/G queue theory.
result Describes propagation of trading signals with finite velocity.
Federated learning optimizes power for reliable V2V communication.
problem Minimizing power consumption for reliable V2V communication.
method Decentralized federated learning for estimating extreme queue lengths.
result Significant reduction in extreme events of queue lengths.
The paper optimizes LLM inference systems through queueing theory.
problem Efficient LLM inference for AI agents under various routing topologies.
method Developed a fluid-limit framework for multi-class batched processing networks under K-FCFS scheduling.
result Proved that work-conserving scheduling algorithms maximize throughput for LLM inference.
Unified model for market dynamics, linking price and order flow.
problem Modeling market dynamics and order flow in a unified framework.
method Markovian market model driven by a hidden Brownian efficient price, signal-driven and queue-reactive models.
result Stability of mid-price around efficient price at macroscopic scale, behavior as diffusion.
In a financial market, for agents with long investment horizons or at times of severe market stress, it is often changes in the asset price that act as the trigger for transactions or shifts in investment position. This suggests the use of price thresholds to simulate agent behavior over much longer timescales than are…
This paper considers a cross-layer adaptive modulation system that is modeled as a Markov decision process (MDP). We study how to utilize the monotonicity of the optimal transmission policy to relieve the computational complexity of dynamic programming (DP). In this system, a scheduler controls the bit rate of the m-qu…
An online learning framework optimizes pricing and capacity in service systems.
problem Optimizing pricing and capacity in dynamic service systems.
method Gradient-based Online Learning in Queue (GOLiQ) framework.
result GOLiQ achieves logarithmic regret bound and improves service provider's performance.
This paper attempts to find out numerically the distribution of the queue-length ratio in the context of a model of preferential attachment. Here we consider two restaurants only and a large number of customers (agents) who come to these restaurants. Each day the same number of agents sequentially arrives and decides w…
The paper improves regret bounds for admission control in queueing systems.
problem Improving regret bounds for admission control in queueing systems.
method Proposes an algorithm inspired by UCRL2 and uses problem structure to bound regret.
result Proves an upper bound on the expected total regret of O ( S log T + m T log T ) O(S\log T + \sqrt{mT \log T}) O ( S log T + m T log T ) . Model for cross-border markets with limited transmission capacities.
problem Limited transmission capacities between two countries' markets.
method Developed a regime-switching process model with high-frequency approximation.
result Analytic tractability allows computation of key market quantities.
Delay Pruning regularizes DyBM for better learning of temporal sequences.
problem Improving DyBM's generalization in learning high-dimensional temporal sequences.
method Delay Pruning: setting some FIFO queue lengths to one with fixed probability.
result Delay Pruning enhances DyBM's performance in learning high-dimensional temporal sequences.
Order submission and cancellation are two constituent actions of stock trading behaviors in order-driven markets. Order submission dynamics has been extensively studied for different markets, while order cancellation dynamics is less understood. There are two positions associated with a cancellation, that is, the price…
Paper improves volatility estimation using a Queue-Reactive model.
problem Volatility estimation from high-frequency data is biased by microstructure noise.
method Uses Queue-Reactive model of limit order book to improve volatility estimation.
result Unified and alternation estimators lead to optimal mean squared error for integrated volatility.
RL optimizes meta-order execution by adapting to market conditions.
problem Optimal execution of large orders while minimizing market impact.
method Data-driven, model-free reinforcement learning with Queue-Reactive Model.
result RL agent learns effective execution policies across various conditions.
The study examines how backrun auctions can protect traders from price manipulation.
problem Price manipulation by arbitrageurs in batched trading venues.
method Developed a laminated queueing model to study price manipulation and introduced a price manipulation coefficient.
result Bound the price manipulation coefficient and found it approximated by a 'zeta value' with measurable parameters.
We examine the dynamics of the bid and ask queues of a limit order book and their relationship with the intensity of trade arrivals. In particular, we study the probability of price movements and trade arrivals as a function of the quote imbalance at the top of the limit order book. We propose a stochastic model in an …
Gittins index optimizes decision-making under uncertainty, even in complex scenarios.
problem Optimal decision-making under uncertainty.
method Gittins index optimizes allocation of resources among uncertain options.
result Gittins index can be effectively applied to practical problems, including Bayesian optimization and queue latency minimization.
Study proves fluid limits of fragmented limit-order markets.
problem Modeling fragmented limit-order markets with small and frequent orders.
method Proved convergence of discrete system to fluid limit characterized by coupled nonlinear ODEs.
result Fluid system converges to stationary equilibrium state over time.
Silent abandonment reduces contact center efficiency by 5%-15%.
problem Measuring customer abandonment and patience in text-based contact centers is challenging due to uncertainty.
method Developed methodologies to identify silent-abandonment customers and estimate customer patience using text analysis and queueing models.
result Silent abandonment accounts for 30%-67% of customer abandonments and reduces system efficiency by 5%-15%.
Proves limiting distributions for Markov chains in random environments.
problem Analyzing Markov chains in random environments.
method Proves existence of limiting distributions using drift and minorization conditions.
result Law of large numbers holds for bounded functionals of the process.
Model shows traders' swarm behavior in markets with subjective predictions.
problem Understanding swarm behavior in markets with traders having subjective market predictions.
method Combination of priority queueing model and mean field theory, nonlinear Markov model.
result Swarm behavior emerges due to traders' reactions to market conditions, not their subjective predictions.
A learning-based algorithm optimizes admission control in a queuing system.
problem Optimizing admission decisions in a queuing system with unknown parameters.
method Proposes a learning-based dispatching algorithm to minimize regret compared to optimal policies.
result Achieves optimal regret bounds for different scenarios of unknown parameters.
Method uses deep learning to estimate traffic intensity.
problem Estimating stochastic intensity of traffic processes.
method Deep neural networks for nonlinear filtering.
result Deep learning method accurately estimates traffic intensity.
New RL algorithm reduces regret in birth-death queueing problems.
problem Efficiency of reinforcement learning in MDPs with large state spaces.
method Modified Ucrl2 algorithm exploiting birth-death structure.
result Regret bound of i l d e O ( E 2 A T ) ilde{\mathcal{O}}(\sqrt{E_2AT}) i l d e O ( E 2 A T ) independent of state space size. Market makers face a trade-off between fill probability and post-fill returns, requiring contrarian strategies.
problem Navigating the trade-off between fill probability and post-fill returns in market making.
method Analysis of live trading data from Binance Bitcoin perpetual.
result A negative correlation between maker fill likelihood and post-fill returns, necessitating contrarian strategies.
New clustering method uses Wasserstein distance to analyze simulation outputs.
problem Analyzing stochastic simulation outputs to uncover relationships and patterns.
method Agglomerative clustering using regularized Wasserstein distance.
result Identifies staffing plans yielding similar performance outcomes.
The paper develops a new model for order book dynamics using Hawkes processes.
problem Capturing the dynamics of order flow and liquidity migration in financial markets.
method Develops a mesoscopic model using Hawkes processes to describe interactions between order arrivals, cancellations, and liquidity movement.
result Derives a diffusive limit for the order book dynamics, providing a unified framework for market microstructure.
Optimal strategy found for liquidating large-tick stocks.
problem Maximizing wealth in liquidating stock positions.
method Semi-Markov decision process, Laplace method, queueing theory, dynamic programming.
result Optimal liquidation policy found.
The paper models CBF dynamics using queueing theory and insurance risk models.
problem Understanding and optimizing the operation of community bail funds.
method Combining queueing theory with classic insurance risk models.
result A fluid limit for the blocking model of CBF operations.
Study of a simplified auction model's price distribution and first passage times in low traffic.
problem Analyzing the price distribution and first passage times in a simplified continuous double auction model.
method Modeling the auction as two M/M/1 queues and studying the low-traffic limit.
result Exact distribution of prices in the low-traffic limit and reasonable approximation for low order arrival rates.