Abstract not found

We propose a method of knowledge reuse for an ensemble of genetic programming-based learners solving a visual learning task. First, we introduce a visual learning method that uses genetic programming individuals to represent hypotheses. Individuals-hypotheses process image representation composed of visual primitives derived from the training images that contain objects to be recognized. The process of recognition is generative, i.e., an individual is supposed to restore the shape of the processed object by drawing its reproduction on a separate canvas. This canonical method is extended with a knowledge reuse mechanism that allows a learner to import genetic material from hypotheses that evolved for the other decision classes (object classes). We compare the performance of the extended approach to the basic method on a real-world tasks of handwritten character recognition, and conclude that knowledge reuse leads to significant convergence speedup and, more importantly, significantly reduces the risk of overfitting.

We consider multitask learning of visual concepts within genetic programming (GP) framework. The proposed method evolves a population of GP individuals, with each of them composed of several GP trees that process visual primitives derived from input images. The two main trees are delegated to solving two different visual tasks and are allowed to share knowledge with each other by calling the remaining GP trees (subfunctions) included in the same individual. The method is applied to the visual learning task of recognizing simple shapes, using generative approach based on visual primitives, introduced in [17]. We compare this approach to a reference method devoid of knowledge sharing, and conclude that in the worst case cross-task learning performs equally well, and in many cases it leads to significant performance improvements in one or both solved tasks.

Abstract. We propose a learning algorithm that reuses knowledge acquired in past learning sessions to improve its performance on a new learning task. The method concerns visual learning and uses genetic programming to represent hypotheses, each of them being a procedure that processes visual primitives derived from the training images. The process of recognition is generative, i.e., a procedure is supposed to restore the shape of the processed object by drawing its reproduction on a separate canvas. This basic method is extended with a knowledge reuse mechanism that allows learners to import genetic material from hypotheses that evolved for the other decision classes (object classes). We compare both methods on a task of handwritten character recognition, and conclude that knowledge reuse leads to signi cant improvement of classi cation accuracy and reduces the risk of over tting. 1

We propose a method of knowledge reuse between evolutionary processes that solve different optimization tasks. We define the method in the framework of tree-based genetic programming (GP) and implement it as code reuse between GP trees that evolve in parallel in separate populations delegated to particular tasks. The technical means of code reuse is a crossbreeding operator which works very similar to standard tree-swapping crossover. We consider two variants of this operator, which differ in the way they handle the incompatibility of terminals between the considered problems. In the experimental part we demonstrate that such code reuse is usually beneficial and leads to success rate improvements when solving the common boolean benchmarks.

We introduce fitnessless coevolution (FC), a novel method of comparative one-population coevolution. FC plays games between individuals to settle tournaments in the selection phase and skips the typical phase of evaluation. The selection operator applies a single-elimination tournament to a randomly drawn group of individuals, and the winner of the final round becomes the result of selection. Therefore, FC does not involve explicit fitness measure. We prove that, under a condition of transitivity of the payoff matrix, the dynamics of FC is identical to that of the traditional evolutionary algorithm. The experimental results, obtained on a diversified group of problems, demonstrate that FC is able to produce solutions that are equally good or better than solutions obtained using fitness-based one-population coevolution with different selection methods.

Abstract. We propose a method of knowledge reuse for an ensemble of genetic programming-based learners solving a visual learning task. First, we introduce a visual learning method that uses genetic programming individuals to represent hypotheses. Individuals-hypotheses process image representation composed of visual primitives derived from given training images that contain objects to be recognized. The process of recognition is generative, i.e., an individual is supposed to restore the shape of the processed object by drawing its reproduction on a separate canvas. This canonical method is in the following extended with a knowledge reuse mechanism that allows a learner to import genetic material from hypotheses that evolved for other decision classes (object classes). We compare the performance of the extended approach to the basic method on a real-world tasks of handwritten character recognition, and conclude that knowledge reuse leads to signi cant convergence speedup and reduces the risk of over tting. 1

Abstract. In this paper we analyze the properties of functional modularity, a concept introduced in [14] for detecting and measuring modularity in problems of automatic program synthesis, in particular by means of genetic programming. The basic components of functional modularity approach are subgoals – entities that embody module’s semantic – and monotonicity, a measure for assessing subgoals ’ potential utility for searching for good modules. For a given subgoal and a sample of solutions decomposed into parts and contexts according to module definition, monotonicity measures the correlation of distance between semantics of solution’s part and the fitness of the solution. The central tenet of this approach is that highly monotonous subgoals can be used to decompose the task and improve search convergence. In the experimental part we investigate the properties of functional modularity using eight instances of problems of Boolean function synthesis. The results show that monotonicity varies depending on problem’s structure of modularity and correctly identifies good subgoals, potentially enabling automatic program decomposition.

In the present position paper, I explore biologicallyinspired computational processes that allow complex high-level collective behaviors to arise from low-level artificial agents (swarmers) – automatically. In contrast to similar projects, I seek elimination of technical constraints that narrow the free development of biologyanalogous behavioral patterns. The result of such swarm evolutions is a fascinating variety of biological, yet completely transparent, analyzable behavior. Results include the spontaneous evolution of an exploration strategy that recently has been mathematically proven to be the optimal one under the conditions given. The work (which is part of my diploma thesis [11]) originally contributes to the field of synthetic biology and the goal was to make evolution milestones in biological swarm collaboration visible. However, I feel that highlevel behavior generation techniques can be migrated to the field of collaborative security and suggest approaches to do so. 1

We introduce potential fitness, a variant of fitness function that operates in the space of schemata and is applicable to tree-based genetic programing. The proposed evaluation algorithm estimates the maximum possible gain in fitness of an individual’s direct offspring. The value of the potential fitness is calculated by analyzing the context semantics and subtree semantics for all contexts (schemata) of the evaluated tree. The key feature of the proposed approach is that a tree is rewarded for the correctly classified fitness cases, but it is not penalized for the incorrectly classified ones, provided that such errors are recoverable by substitution of an appropriate subtree (which is however not explicitly considered by the algorithm). The experimental evaluation on a set of seven boolean benchmarks shows that the use of potential fitness may lead to better convergence and higher success rate of the evolutionary run.