Genetic programming for cross-task knowledge sharing

Abstract

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.

Automatic generation and exploitation of related problems in genetic programming

“…Let us finally note that the above algorithm shares some common elements with our previous studies on cross-task knowledge reuse [4] and knowledge reuse for visual learning [5], where we have demonstrated that crossing over individuals that solve different visual tasks speeds up the learning. Here, however, we are interested in a scenario where the input to the method is a single task (problem).…”

Generative learning of visual concepts using multiobjective genetic programming

“…This feature potentially enables further re-use of acquired knowledge in other tasks; such possibility has been already conrmed in a preliminary study [28].…”

Evolving strategy for a probabilistic game of imperfect information using genetic programming

“…There is no better way to appreciate the effectiveness of the simulated evolution than through a personal experience. Check at [11] if you can beat BrilliAnt.…”

“…Even if we assume that inexperienced beginners account for great part of this statistics, these figures clearly indicate that the strategy elaborated by our approach is challenging for humans. The reader is encouraged to visit the Web page [11] and measure swords with BrilliAnt.…”