Abstract:Understanding how players adjust their strategies in games, based on their experience, is a crucial tool for policymakers. It enables them to forecast the system's eventual behavior, exert control over the system, and evaluate counterfactual scenarios. The task becomes increasingly difficult when only a limited number of observations are available or difficult to acquire. In this work, we introduce the Side-Information Assisted Regression (SIAR) framework, designed to identify game dynamics in multiplayer normal-form games only using data from a short run of a single system trajectory. To enhance system recovery in the face of scarce data, we integrate side-information constraints into SIAR, which restrict the set of feasible solutions to those satisfying game-theoretic properties and common assumptions about strategic interactions. SIAR is solved using sum-of-squares (SOS) optimization, resulting in a hierarchy of approximations that provably converge to the true dynamics of the system. We showcase that the SIAR framework accurately predicts player behavior across a spectrum of normal-form games, widely-known families of game dynamics, and strong benchmarks, even if the unknown system is chaotic.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to identify dynamic behaviors in multi - player normal - form games through limited observational data in the case of data scarcity. Specifically, the researchers hope to develop a method that can accurately predict players' behavior changes using only short - time running data from a single system trajectory. This problem is particularly important in practical applications because in many cases, obtaining a large amount of data may be difficult or expensive, such as short - term behavior prediction in complex systems like traffic networks and financial markets. ### Core Challenges of the Problem 1. **Data Scarcity**: Traditional methods usually rely on a large amount of observational data to identify system dynamics, but in many real - world scenarios, obtaining a large amount of data is unrealistic. 2. **Real - Time Requirements**: In some application scenarios, decision - makers need to make effective decisions based on limited data within a short time. 3. **Complexity and Chaotic Behavior**: Some game dynamics may exhibit complex or even chaotic behaviors, which makes traditional equilibrium - point - based methods difficult to apply. ### Overview of the Solution To solve the above problems, the authors introduce the **Side - Information Assisted Regression (SIAR)** framework. The main features of this framework include: 1. **Utilizing Side - Information**: By introducing theoretical properties and common assumptions in game theory as constraints, the space of feasible solutions is restricted, thereby improving the accuracy of the model in the case of data scarcity. 2. **Sum - of - Squares Optimization (SOS)**: Use SOS optimization techniques to solve regression problems, ensuring that the found solutions have theoretical approximation guarantees. 3. **Polynomial Regression**: Approximate players' behavior update strategies through polynomial regression, enabling the model to perform effective parameter estimation under limited data conditions. ### Specific Application Scenarios The SIAR framework has demonstrated its effectiveness in several classic benchmark tests, including but not limited to: - **Periodic Replicator Dynamics**: Such as the dynamic behavior in the Matching Pennies game. - **Chaotic Replicator Dynamics**: Such as the dynamic behavior in the ε - perturbed Rock - Paper - Scissors game. - **Equilibrium Selection Prediction in Multi - Player Atomic Congestion Games**: Games involving up to hundreds of pure - attracting Nash equilibria. ### Summary This paper aims to solve the problem of identifying multi - player game dynamics in the case of data scarcity. It proposes the SIAR framework, which combines the theoretical properties of game theory and SOS optimization techniques and shows good performance and applicability in various complex scenarios. This method not only helps to understand players' behavior patterns but also provides a powerful tool for policymakers to make wiser decisions under limited data conditions.

Data-Scarce Identification of Game Dynamics via Sum-of-Squares Optimization

Learning and steering game dynamics towards desirable outcomes

Sparse Identification of Nonlinear Dynamics with Side Information (SINDy-SI)

Network Learning from Best-Response Dynamics in LQ Games.

Inferring Short-Sightedness in Dynamic Noncooperative Games

Inducing Social Optimality in Games via Adaptive Incentive Design

VISER: A Tractable Solution Concept for Games with Information Asymmetry

Compressive-sensing-assisted mixed integer optimization for dynamical system discovery with highly noisy data

Information Compression in Dynamic Games

Inferring Objectives in Continuous Dynamic Games from Noise-Corrupted Partial State Observations

Identifying Nonlinear Dynamics with High Confidence from Sparse Data

Learning sparse nonlinear dynamics via mixed-integer optimization

Learning Probably Approximately Correct Maximin Strategies in Simulation-Based Games with Infinite Strategy Spaces

Generation of probabilistic synthetic data for serious games: A case study on cyberbullying

Learning to Control Unknown Strongly Monotone Games

Subsidy design for better social outcomes

Gamifying optimization: a Wasserstein distance-based analysis of human search

A Risk-Averse Equilibrium for Multi-Agent Systems

Accurate Data-Driven Surrogates of Dynamical Systems for Forward Propagation of Uncertainty

Data-Driven Behaviour Estimation in Parametric Games