Abstract:Chance-constrained problems involve stochastic components in the constraints which can be violated with a small probability. We investigate the impact of different types of chance constraints on the performance of iterative search algorithms and study the classical maximum coverage problem in graphs with chance constraints. Our goal is to evolve reliable chance constraint settings for a given graph where the performance of algorithms differs significantly not just in expectation but with high confidence. This allows to better learn and understand how different types of algorithms can deal with different types of constraint settings and supports automatic algorithm selection. We develop an evolutionary algorithm that provides sets of chance constraints that differentiate the performance of two stochastic search algorithms with high confidence. We initially use traditional approximation ratio as the fitness function of (1+1)~EA to evolve instances, which shows inadequacy to generate reliable instances. To address this issue, we introduce a new measure to calculate the performance difference for two algorithms, which considers variances of performance ratios. Our experiments show that our approach is highly successful in solving the instability issue of the performance ratios and leads to evolving reliable sets of chance constraints with significantly different performance for various types of algorithms.

What problem does this paper attempt to address?

### What problem does this paper attempt to solve? This paper aims to solve the problem of algorithm selection and performance differences in the Chance - constrained Maximum Coverage Problem (CC - MCP). Specifically, the author focuses on how to generate reliable discriminatory instances that can significantly distinguish the performance of different random search algorithms when solving CC - MCP. #### Background and Motivation 1. **Chance - constrained optimization problems**: Chance - constrained optimization problems contain random components that can be violated with a very small probability. Such problems are very common in the real world because many practical problems have uncertainties. 2. **Algorithm selection problem**: Different algorithms exhibit different performances when dealing with different types of problems. Therefore, it is very important to understand which algorithms are most suitable for specific types of problems. This involves learning the relationship between algorithm performance and problem characteristics through a large amount of experimental data. 3. **Importance of benchmarking**: In order to evaluate and compare the performance of different algorithms, a set of benchmark problems with different complexities is required. Especially for chance - constrained optimization problems, the existing benchmark instances often fail to well show the performance differences between different algorithms. #### Research Objectives The author's goal is to develop an evolutionary algorithm for generating reliable discriminatory instances that can: - **Significantly distinguish the performance of two random search algorithms**: Make one algorithm easy to solve and the other difficult to solve. - **Have high confidence and low variance**: Ensure that the generated instances are stable in multiple runs and reduce the impact of uncertainty. #### Method Overview 1. **Deficiencies of traditional methods**: Initially, the author used the traditional approximation ratio as a fitness function to generate discriminatory instances, but this method was unstable in the chance - constrained setting and the generated instances had low reliability. 2. **Introduction of a new performance measure**: To solve the above problems, the author introduced a new performance measurement method - the Discounted Expected Approximation Ratio. This method takes into account the variance of the performance ratio in each run, thereby improving the stability of the results. 3. **Design of the evolutionary algorithm**: The author designed a (1 + 1) evolutionary algorithm (EA) to generate discriminatory instances by adjusting the expected cost and variance of each node. The algorithm uses an elitist strategy to select the best individuals and evolves more reliable instances by maximizing the performance difference. #### Experimental Verification The author verified the effectiveness of the new method through a series of experiments, and the results showed that the new method can generate more stable discriminatory instances and significantly improve the reliability of algorithm selection. ### Summary The main contribution of this paper is to propose a new evolutionary algorithm framework for generating reliable discriminatory instances to better understand and evaluate the performance of different algorithms when solving the Chance - constrained Maximum Coverage Problem. This method not only improves the reliability of benchmarking but also provides strong support for future algorithm selection and optimization research.

Evolving Reliable Differentiating Constraints for the Chance-constrained Maximum Coverage Problem

Runtime Analysis of Single- and Multi-Objective Evolutionary Algorithms for Chance Constrained Optimization Problems with Normally Distributed Random Variables

A Relative Feasibility Degree Based Approach for Constrained Optimization Problems.

Runtime Performance of Evolutionary Algorithms for the Chance-constrained Makespan Scheduling Problem

Optimizing Chance-Constrained Submodular Problems with Variable Uncertainties

Optimizing Monotone Chance-Constrained Submodular Functions Using Evolutionary Multi-Objective Algorithms

Using 3-Objective Evolutionary Algorithms for the Dynamic Chance Constrained Knapsack Problem

Adaptive Partitioning for Chance-Constrained Problems with Finite Support

Runtime Analysis of RLS and the (1+1) EA for the Chance-constrained Knapsack Problem with Correlated Uniform Weights

Sampling-based Pareto Optimization for Chance-constrained Monotone Submodular Problems

A genetic algorithm for chance constrained programming

Analyzing design principles for competitive evolution strategies in constrained search spaces

Chance Constrained Stochastic Optimal Control Based on Sample Statistics With Almost Surely Probabilistic Guarantees

Exploiting Overlap Information in Chance-constrained Program with Random Right-hand Side

Chance Constrained Selection of the Best.

An adaptive helper and equivalent objective evolution strategy for constrained optimization

Chance Constrained Probability Measure Optimization: Problem Formulation, Equivalent Reduction, and Sample-based Approximation

An Enhanced Adaptive Differential Evolution Approach for Constrained Optimization Problems

An Empirical Quantile Estimation Approach to Nonlinear Optimization Problems with Chance Constraints

Nonconvex and Nonsmooth Approaches for Affine Chance-Constrained Stochastic Programs

Balance Programming Between Target And Chance With Application In Building Optimal Bidding Strategies For Generation Companies