This work presents a novel learning method in the context of embodied artificial intelligence and self-organization, which has as few assumptions and restrictions as possible about the world and the underlying model. The learning rule is derived from the principle of maximizing the predictive information in the sensorimotor loop. It is evaluated on robot chains of varying length with individually controlled, non-communicating segments. The comparison of the results shows that maximizing the predictive information per wheel leads to a higher coordinated behavior of the physically connected robots compared to a maximization per robot. Another focus of this paper is the analysis of the effect of the robot chain length on the overall behavior of the robots. It will be shown that longer chains with less capable controllers outperform those of shorter length and more complex controllers. The reason is found and discussed in the information-geometric interpretation of the learning process.
The field of embodied intelligence emphasises the importance of the morphology and environment with respect to the behaviour of a cognitive system. The contribution of the morphology to the behaviour, commonly known as morphological computation, is well-recognised in this community. We believe that the field would benefit from a formalisation of this concept as we would like to ask how much the morphology and the environment contribute to an embodied agent's behaviour, or how an embodied agent can maximise the exploitation of its morphology within its environment. In this work we derive two concepts of measuring morphological computation, and we discuss their relation to the Information Bottleneck Method. The first concepts asks how much the world contributes to the overall behaviour and the second concept asks how much the agent's action contributes to a behaviour. Various measures are derived from the concepts and validated in two experiments that highlight their strengths and weaknesses.
Abstract. Using discrete-time dynamics of a two neuron network with recurrent connectivity it is shown that for specific parameter configurations the output signals of neurons can be of almost sinusoidal shape. These networks live near the Sacker-Neimark bifurcation set, and are termed SO(2)-networks, because their weight matrices correspond to rotations in the plane. The discretized sinus-shaped waveform is due to the existence of quasi-periodic attractors. It is shown that the frequency of the oscillators can be controlled by only one parameter. Signals from the neurons have a phase shift of π/2 and may be useful for various kinds of applications; for instance controlling the gait of legged robots.
In recent years, the application of information theory to the field of embodied intelligence has turned out to be extremely fruitful. Here, several measures of information flow through the sensorimotor loop of an agent are of particular interest. There are mainly two ways to apply information theory to the sensorimotor setting.First, information-theoretic measures can be used within various analysis methods. Sensorimotor interactions of an embodied agent lead to the emergence of redundancy and structure of the agent's intrinsic processes. Understanding the generation of structure in the sensorimotor process and its exploitation is important within the field of embodied intelligence (Pfeifer and Bongard 2006). The quantification and analysis of information flows through an agent's sensorimotor loop from the perspective of an external observer, that is from the perspective of a scientist, proves to be effective in this regard Sporns 2005, 2006). Here, transfer entropy (Schreiber 2000) has been used in order to quantify the flows of information between various processes of the sensorimotor loop, such as the sensor process on the actuator process. Furthermore, excess entropy, also known as predictive information (Bialek et al. 2001), has been used to analyse the interplay between information-theoretic measures and behavioral patterns of embodied agents .Second, information-theoretic measures can be used as objective functions for self-organized learning. This is based on the hypothesis that learning in natural intelligent systems is partly governed by an information-theoretic optimisation principle. Corresponding studies aim at the implementation of related principles, so-called
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.