Other Evolutionary Algorithms

Several other types of evolutionary algorithms (EAs) exist, differing in the selection method for parents and the way in which one population is formed from the previous one or in the interpretation of the GA strings. In addition, evolutionary agent-based models are starting to appear. Although these methods show promise, they are currently less widely used in science than the GA, so are covered here briefly. 

---

As with all other evolutionary algorithms, the initial population is generated at random. The exact manner in which this happens, however, is entirely dependent on the problem domain. EP can handle any structure, whether as simple as the binary strings used by GAs, sets of real numbers, or even highly complex structures like neural networks. 

Once the population has been generated, each individual is tested, again in the same manner as other evolutionary algorithms. Another difference occurs when generating the next generation. Each individual in the population is looked upon as a species, and if that individual is chosen to survive into the next generation, the species should evolve, in that, although it is somewhat different from its predecessor, there should still be some clear behavioral similarities. 

Population sizes in ES are extremely small compared to other evolutionary algorithms. At the extreme, it is possible to have a population size of just one. This is known as a two-membered, or (1 -p 1), ES. There are considered to be two members because there is always a parent and a child. The (1 -P 1) refers to the fact that there is one parent, which produces one offspring in each generation. In this case, there is no crossover, just mutation. When the child is created, it is compared to the parent and, if it has a greater fitness, replaces the parent. 

Zitzler and Thiele (1999) proposed the strength Pareto evolutionary algorithm (SPEA) with combination of several features of midti-objective evolutionary algorithms in a imique manner. The SPEA has some similarities in its process to other evolutionary algorithms (Zitzler and Thiele 1999) in that it... 

After the candidate downstream gene modules are selected by GA, FSO is proposed to determine the parameters in the NN model. Particle swarm optimization is motivated by the behavior of bird flocking or fish blocking, originally intended to explore optimal or near-optimal solutions in sophisticated continuous spaces (Kennedy and Eberhart 1995). Its main difference from other evolutionary algorithms (e.g., GA) is that PSO relies on cooperation rather than competition. Good solutions in the problem set are shared with their less-fit ones so that the entire population improves. 

In common with most other AI algorithms, the GA contains several variables whose values are chosen at the start of a run. Decisions must also be made about how to implement the evolutionary operators within the algorithm because there may be more than one way in which the operators can be used. We shall deal with the permissible values of these parameters and the factors that help us to choose among the evolutionary operators as they are introduced. 

By a comparison of the new evolutionary algorithm s performance with state-of-the-art solvers for a real-world scheduling problem it was found that the new algorithm shows a competitive performance. In contrast to the other algorithms the evolutionary algorithm was able to provide relatively good solutions in short computation times. 

An evolutionary algorithm is included in the current release of Frontline Systems Premium Excel Solver (for current information, see www.frontsys.com). It is invoked by choosing Standard Evolutionary from the Solver dropdown list in the Solver Parameters Dialog Box. The other nonlinear solver is Standard GRG Nonlinear, which is the GRG2 solver described in Section 8.7. As discussed there, GRG2 is a gradient-based local solver, which will find the nearest local solution to its starting point. The evolutionary solver is much less likely to stop at a local minimum, as we illustrate shortly. 

Evolutionary algorithms have been widely used in other areas of physical chemistry, such as photonics. An interesting application is from Lipson et al.16 where the spontaneous emergence of structure was evident when using GAs to design a high-confinement photonic structure. 

Evolutionary algorithms are frequently used to find optimal solutions in many different problem areas. They are based on Darwin s principle of survival of the fittest (Darwin 1996 Maynard Smith 1993). A population of individuals has to compete with other individuals for access to food and mates. Only the successful ones are allowed to reproduce. This leads to the reproduction of certain inheritable traits into the next generation. 

A detailed description of various kinds of genetic algorithms, as well as of other kinds of the broader class of evolutionary algorithms, can be found in specific monographs [23-34], Here, only those particular features of genetic algorithms... 

WGS 1) while others were revealed when using a DoE step for the initial design of the first library and using statistical analyses of the data set (WGS 2) and vice versa. This stresses that the choice of evolutionary algorithms partially determines the type of discovered formulas. 

The neural network architecture optimized by the evolutionary algorithm could be analyzed for a biochemical interpretation and feature extraction. One may infer the importance of input properties based on the relative connectivity of the input units. For example, bulkiness, which was not connected at all, was probably unimportant. On the other hand, units for polarity, refractivity, hydrophobicity, and surface area were highly connected, indicating these are important features of membrane transition regions. 

Several other search techniques in very large virtual spaces are available, in particular optimization by means of evolutionary algorithms (genetic algorithms... 

---