The goal of getting computers to automatically solve problems is central to artificial intelligence, machine learning, and the broad area encompassed by what Turing called “machine intelligence ” [161, 162].

We present an adaptation of the standard genetic program (GP) to hierarchically decomposable, multi-agent learning problems. To break down a problem that requires cooperation of multiple agents, we use the team objective function to derive a simpler, intermediate objective function for pairs of cooperating agents. We apply GP to optimize first for the intermediate, then for the team objective function, using the final population from the earlier GP as the initial seed population for the next. This layered learning approach facilitates the discovery of primitive behaviors that can be reused and adapted towards complex objectives based on a shared team goal.

Although fuzzy systems demonstrated their ability to solve different kinds of problems in various applications, there is an increasing interest on augmenting them with learning capabilities. Two of the most successful approaches to hybridise fuzzy systems with adaptation methods have been made in the realm of soft computing: neuro-fuzzy systems and genetic fuzzy systems hybridise the approximate reasoning method of fuzzy systems with the learning capabilities of neural networks and evolutionary algorithms. This contribution focus on genetic fuzzy systems, paying special attention to genetic fuzzy rule based systems, giving a brief overview of the field.

The paper discusses three major issues. First, it discusses why it makes sense to approach problems in a hierarchical fashion. It defines the class of hierarchically decomposable functions that can be used to test the algorithms that approach problems in this fashion. Finally, the Bayesian optimization algorithm (BOA) is extended in order to solve the proposed class of problems.

Abstract. Most of the existing work in the study of bargaining behavior uses techniques from game theory. Game theoretic models for bargaining assume that players are perfectly rational and that this rationality is common knowledge. However, the perfect rationality assumption does not hold for real-life bargaining scenarios with humans as players, since results from experimental economics show that humans find their way to the best strategy through trial and error, and not typically by means of rational deliberation. Such players are said to be boundedly rational. In playing a game against an opponent with bounded rationality, the most effective strategy of a player is not the equilibrium strategy but the one that is the best reply to the opponent’s strategy. The evolutionary model provides a means for studying the bargaining behaviour of boundedly rational players. This paper provides a comprehensive comparison of the game theoretic and evolutionary approaches to bargaining by examining their assumptions, goals, and limitations. We then study the implications of these differences from the perspective of the software agent developer.

Digital sound synthesizers, ubiquitous today in sound cards, software and dedicated hardware, use algorithms (Sound Synthesis Techniques, SSTs) capable of generating sounds similar to those of acoustic instruments and even totally novel sounds. The design of SSTs is a very hard problem. It is usually assumed that it requires human ingenuity to design an algorithm suitable for synthesizing a sound with certain characteristics. Many of the SSTs commonly used are the fruit of experimentation and a long refinement processes. A SST is determined by its “functional form ” and “internal parameters”. Design of SSTs is usually done by selecting a fixed functional form from a handful of commonly used SSTs, and performing a parameter estimation technique to find a set of internal parameters that will best emulate the target sound. A new approach for automating the design of SSTs is proposed. It uses a set of examples of the desired behavior of the SST in the form of “inputs + target sound”. The approach is capable of suggesting novel functional forms and their internal parameters, suited to follow closely the given examples.

This paper describes a biologically inspired domain-independent technique, called genetic programming, that automatically creates computer programs to solve problems. We argue that the field of design is a useful testbed for determining whether an automated technique can produce results that are competitive with human-produced results. We present several results that are competitive with the products of human creativity and inventiveness. This claim is supported by the fact that each of the results infringe on previously issued patents. This paper presents a candidate set of criteria that identify when a machine-created solution to a problem is competitive with a human-produced result. 1.

Genetic programming is a means of automatically evolving programs to perform a particular task or solve a particular problem using the Darwinian principle of survival of the fittest. Many genetic programming problems can suffer from early convergence, that is, the genetic programming run terminates before the optimal program has evolved. Early convergence is a hindrance to genetic programming especially for problems which need signi cant amounts of computing time. This project describes a method of preventing early convergence by replacing similar programs. A percentage of the most similar programs are replaced by randomly generated programs. This method uses the number of changes reported by the UNIX program diff to estimate how similar a program is to the rest of the population. We performed experiments using no replacement and replacement on the MAX problem, a problem known to suffer from early convergence, and the Robocup simulator league domain. Using a replacement rate of 10% in the MAX domain, increased the success rate from 16% (using no replacement) to 42%. Performing similarity replacement in the Robocup domain increased the number of runs which obtained successful players, from 2 out of the 5 runs using no replacement, to 4 out of the 5 runs using 10% replacement. The quality of the players in the successful runs was also improved. Performing replacement every 2nd, 5th, or 10th generation did not significantly reduce the number of successful runs in the MAX domain when using a replacement rate of 10%. Replacing a percentage of the most similar programs prevented early convergence more often than when no replacement was used. Our results suggest that performing similarity replacement is worthwhile in problems where the cost of computing the t...

Abstract. Loops are rarely used in genetic programming (GP), because they lead to massive computation due to the increase in the size of the search space. We have investigated the use of loops with restricted semantics for a problem in which there are natural repetitive elements, that of distinguishing two classes of images. Using our formulation, programs with loops were successfully evolved and performed much better than programs without loops. Our results suggest that loops can successfully used in genetic programming in situations where domain knowledge is available to provide some restrictions on loop semantics. 1

Abstract not found