An optimization procedure using empirical models as an approximation of expensive functions is presented. The model is trained on the current set of evaluated solutions and can be used to search for promising solutions. These solutions are then evaluated on the expensive function. The resulting iterative procedure is analyzed and compared on test problems and shows fast and robust convergence. ...

Self-adaptive mutations are known to endow evolutionary algorithms (EAs) with the ability of locating local optima quickly and accurately, whereas it was unknown whether these local optima are finally global optima provided that the EA runs long enough. In order to answer this question it is assumed that the -EA with self-adaptation is located in the vicinity of a local solution with objective ...

Evolutionary Algorithms (EA) usually carry out an efficient exploration of the search-space, but get often trapped in local minima and do not prove the optimality of the solution. Interval-based techniques, on the other hand, yield a numerical proof of optimality of the solution. However, they may fail to converge within a reasonable time due to their inability to quickly compute a good approxi...

Monte-Carlo localization uses particle filtering to estimate the position of the robot. The method is known to suffer from the loss of potential positions when there is ambiguity present in the environment. Since many indoor environments are highly symmetric, this problem of premature convergence is problematic for indoor robot navigation. It is, however, rarely studied in particle filters. We ...

The optimization by genetic algorithms often comes along with premature convergence bias, especially in the multimodal problems. In the paper, we propose and test two mechanisms to avoid the premature convergence of genetic algorithms by preserving the population diversity in two different manners. These are the dynamic application of many genetic operators, based on the average progress, and t...

Genetic Algorithms (GAs), rst proposed by John Holland in the early seventies, are growing in stature as tools in the elds of machine learning and function optimization. GAs model evolution of life. To solve a particular task, a genetic algorithm creates and maintains a population of organisms, probabilistically modifying the population , while seeking a near-optimal solution to the task at han...

Pattern search learning is known for simplicity and faster convergence. However, one of the downfalls of this learning is the premature convergence problem. In this paper, we show how we can avoid the possibility of being trapped in a local pit by the introduction of stochastic value. This improved pattern search is then applied on a recurrent type neuro-fuzzy network (ANFIS) to solve time seri...

A rapid genetic algorithm based on chaos mechanism is presented in this paper. We introduced the chaos mechanism into the genetic algorithm to remedy the defect of premature convergence in the genetic algorithm, then continuously compressed the searching intervals of the optimization variable for increasing convergence speed. Experiments indicate that this method is a rapid and effective evolut...

The notion of using a meta-heuristic approach to solve nonlinear resource-leveling problems has been intensively studied in recent years. Premature convergence and poor exploitation are the main obstacles for the heuristic algorithms. Analyzing the characteristics of the project topology network, this paper introduces a directional ant colony optimization (DACO) algorithm for solving nonlinear ...

The Travelling Salesmen Problem (TSP) is one of the most important and famous combinational optimization problems that aim to find the shortest tour. In this problem, the salesman starts to move from an arbitrary place called depot and after visiting all nodes, finally comes back to depot. Solving this problem seems hard because program statement is simple and leads this problem belonging to NP...