Christian Horoba

Multiplicative approximations and the hypervolume indicator

Indicator-based algorithms have become a very popular approach to solve multi-objective optimization problems. In this paper, we contribute to the theoretical understanding of algorithms maximizing the hypervolume for a given problem by distributing μ points on the Pareto front. We examine this common approach with respect to the achieved multiplicative approximation ratio for a given multi-objective problem and relate it to a set of μ points on the Pareto front that achieves the best possible approximation ratio. For the class of linear fronts and a class of concave fronts, we prove that the hypervolume gives the best possible approximation ratio. In addition, we examine Pareto fronts of different shapes by numerical calculations and show that the approximation computed by the hypervolume may differ from the optimal approximation ratio.

Benefits and drawbacks for the use of epsilon-dominance in evolutionary multi-objective optimization

Using diversity mechanisms in evolutionary algorithms for multi-objective optimization problems is considered as an important issue for the design of successful algorithms. This is in particular the case for problems where the number of non-dominated feasible objective vectors is exponential with respect to the problem size. In this case the goal is to compute a good approximation of the Pareto front. We investigate how this goal can be achieved by using the diversity mechanism of epsilon-dominance and point out where this concept is provably helpful to obtain a good approximation of an exponentially large Pareto front in expected polynomial time. Afterwards, we consider the drawbacks of this approach and point out situations where the use of epsilon-dominance slows down the optimization process significantly.

Running Time Analysis of ACO Systems for Shortest Path Problems

Ant Colony Optimization (ACO) is inspired by the ability of ant colonies to find shortest paths between their nest and a food source. We analyze the running time of different ACO systems for shortest path problems. First, we improve running time bounds by Attiratanasunthron and Fakcharoenphol [Information Processing Letters , 105(3):88---92, 2008] for single-destination shortest paths and extend their results for acyclic graphs to arbitrary graphs. Our upper bound is asymptotically tight for large evaporation factors, holds with high probability, and transfers to the all-pairs shortest paths problem. There, a simple mechanism for exchanging information between ants with different destinations yields a significant improvement. Our results indicate that ACO is the best known metaheuristic for the all-pairs shortest paths problem.

Evolutionary algorithms and dynamic programming

Recently, it has been proven that evolutionary algorithms produce good results for a wide range of combinatorial optimization problems. Some of the considered problems are tackled by evolutionary algorithms that use a representation, which enables them to construct solutions in a dynamic programming fashion. We take a general approach and relate the construction of such algorithms to the development of algorithms using dynamic programming techniques. Thereby, we give general guidelines on how to develop evolutionary algorithms that have the additional ability of carrying out dynamic programming steps.

Analysis of a simple evolutionary algorithm for the multiobjective shortest path problem

We present a natural fitness function f for the multiobjective shortest path problem, which is a fundamental multiobjective combinatorial optimization problem known to be NP-hard. Thereafter, we conduct a rigorous runtime analysis of a simple evolutionary algorithm (EA) optimizing f. Interestingly, this simple general algorithm is a fully polynomial-time randomized approximation scheme (FPRAS) for the problem under consideration, which exemplifies how EAs are able to find good approximate solutions for hard problems.

Ant colony optimization for stochastic shortest path problems

We consider Ant Colony Optimization (ACO) for stochastic shortest path problems where edge weights are subject to noise that reflects delays and uncertainty. The question is whether the ants can find or approximate shortest paths in the presence of noise. We first prove a general upper bound for the time until the algorithm finds an approximation for arbitrary, independent noise values. For independent gamma-distributed noise we prove lower bounds for the time until a good approximation is found. We construct a graph where the ants cannot find a reasonable approximation, even in exponential time. The last result changes when the noise is perfectly correlated as then the ants find shortest paths efficiently.

Exploring the runtime of an evolutionary algorithm for the multi-objective shortest path problem**

We present a natural vector-valued fitness function f for the multi-objective shortest path problem, which is a fundamental multi-objective combinatorial optimization problem known to be NP-hard. Thereafter, we conduct a rigorous runtime analysis of a simple evolutionary algorithm (EA) optimizing f. Interestingly, this simple general algorithm is a fully polynomial-time randomized approximation scheme (FPRAS) for the problem under consideration, which exemplifies how EAs are able to find good approximate solutions for hard problems. Furthermore, we present lower bounds for the worst-case optimization time.

Maximal age in randomized search heuristics with aging

The concept of aging has been introduced and applied in many different variants in many different randomized search heuristics. The most important parameter is the maximal age of search points. Considering static pure aging known from artificial immune systems in the context of simple evolutionary algorithms, it is demonstrated that the choice of this parameter is both, crucial for the performance and difficult to set appropriately. The results are derived in a rigorous fashion and given as theorems with formal proofs. An additional contribution is the presentation of a general method to combine fitness functions into a function with stronger properties than its components. By application of this method we combine a function where the maximal age needs to be sufficiently large with a function where the maximal age needs to be sufficiently small. This yields a function where an appropriate age lies within a very narrow range.

Illustration of fairness in evolutionary multi-objective optimization

It is widely assumed that evolutionary algorithms for multi-objective optimization problems should use certain mechanisms to achieve a good spread over the Pareto front. In this paper, we examine such mechanisms from a theoretical point of view and analyze simple algorithms incorporating the concept of fairness. This mechanism tries to balance the number of offspring of all individuals in the current population. We rigorously analyze the runtime behavior of different fairness mechanisms and present illustrative examples to point out situations, where the right mechanism can speed up the optimization process significantly. We also indicate drawbacks for the use of fairness by presenting instances, where the optimization process is slowed down drastically.

Approximating Pareto-Optimal Sets Using Diversity Strategies in Evolutionary Multi-Objective Optimization

Often the Pareto front of a multi-objective optimization problem grows exponentially with the problem size. In this case, it is not possible to compute the whole Pareto front efficiently and one is interested in good approximations. We consider how evolutionary algorithms can achieve such an approximation by using different diversity mechanisms. We discuss some well-known approaches such as the density estimator and the e -dominance approach and point out when and how such mechanisms provably help to obtain a good approximation of the Pareto-optimal set.