Carlos H. C. Ribeiro

Dynamic Behaviour of Chaotic Cellular Automata - A Comparative Entropy Analysis of Regular Lattices and Small-World Structures

This work investigates elementary cellular automata comparing their dynamic evolution on two topologies: the usual regular lattice and a small-world network constructed as a directed graph. The number of iterations for any initial perturbation to propagate over all cells of the automaton corresponds to how quickly an initial configuration reaches a high entropy state, and the temporal rate of propagation of the perturbation can be evaluated from a chaotic set of elementary cellular automata rules using an entropy measure. Results indicate that the average entropy can be nearly tripled in the small-world network topology, suggesting much faster applications (e.g. in Cryptography) from alterations in the topological arrangement of usual cellular automata regular lattices.

An attentive multi-sensor based system for mobile robotics

Usually embodied in real environments, robots are expected to perceive and act in their surrounding world in a human-like fashion [1], through perception, reasoning, planning and decision making processes. Although high computational power is available nowadays, the complexity of real scenes - continuous, partially unknown and usually unpredictable - can not be taken for granted. To actively act in this environment, robots are fully equipped with an enormous amount of sensors, overwhelming them with an immense amount of data that can not fully be processed and that constantly changes across time and space. To overcome this problem, the natural human filter - Attention - could be used as inspiration. This paper proposes an architecture that supports a transfer of domain from visual models to a robotics domain that uses sensors such as range scanners and sonars. Furthermore, it discusses the possibility of using multiple sensors to define multiple features. The experiments were performed in a simulated high fidelity environment and results have shown that the model proposed can account for detecting salient stimuli according to the modeled features.

An analysis of feature-based and state-based representations for module-based learning in mobile robots

The information available to robots in real tasks is widely distributed both in time and space, requiring the agent to search for relevant information. In this paper, we implement a solution that uses qualitative and quantitative knowledge to turn robot tasks able to be treated by reinforcement learning (RL) algorithms. The steps of this procedure include: 1) to decompose the overall task into smaller ones, using abstraction and macro-operators, thus achieving a discrete action space; 2) to use observation functions of the environment - here called features - to achieve both time and state space discretisation; 3) to use quantitative knowledge to design controllers that are able to solve the subtasks; 4) to learn the coordination of these behaviours using RL, more specifically Q-learning. The approach was verified on an increasingly complex set of robot tasks using a Khepera robot simulator. Two approaches for space discretisation were used, one based on features and the other on states. The learned policies over these two models were compared to a predefined hand-crafted one. It was found that the learned policy over the state-based discretisation leads quickly to good results, although it can not be applied to complex tasks, where the state space representation becomes computationally unfeasible.

CONAIM: A Conscious Attention-Based Integrated Model for Human-Like Robots

Understanding consciousness is one of the most fascinating challenges of our time. From ancient civilizations to modern philosophers, questions have been asked on how one is conscious of his/her own existence and about the world that surrounds him/her. Although there is no precise definition for consciousness, there is an agreement that it is strongly related to human cognitive processes such as attention, a process capable of promoting a selection of a few stimuli from a huge amount of information that reaches us constantly. In order to bring the consciousness discussion to a computational scenario, this paper presents conscious attention-based integrated model (CONAIM), a formal model for machine consciousness based on an attentional schema for human-like agent cognition that integrates: short- and long-term memories, reasoning, planning, emotion, decision-making, learning, motivation, and volition. Experimental results in a mobile robotics domain show that the agent can attentively use motivation, volition, and memories to set its goals and learn new concepts and procedures based on exogenous and endogenous stimuli. By performing computation over an attentional space, the model also allowed the agent to learn over a much reduced state space. Further implementation under this model could potentially allow the agent to express sentience, self-awareness, self-consciousness, autonoetic consciousness, mineness, and perspectivalness.

A Comparative Study between the Dynamic Behaviours of Standard Cellular Automata and Network Cellular Automata Applied to Cryptography

The dynamical behavior of cellular automata (CA) transition rules are objects of study for different knowledge fields. This paper is about the development of cryptographic methods using CA‐like transition rules. We investigate the dynamic behavior of rules that are not able to propagate a perturbation inserted in the initial lattice considering variations of the regular connection structure of CAs that are akin to the small‐word network construction process. Extensive experimental results indicate that such modifications in the CA connection structure will produce large changes in the dynamic behavior of the evaluated rules, suggesting that it is possible to increase considerably the space of possible cryptographic keys to methods based on CA rules, provided a different topological construct for the CA lattice is considered.

Solving Problems with Extended Reachability Goals through Reinforcement Learning on Propositionally Constrained State Spaces

Finding a near-optimal action policy towards a goal state can be a complex task for intelligent autonomous agents, especially in a model-free environment with unknown rewards and under state space constraints. In such a situation, it is not possible to plan ahead which is the best action to execute at each moment, and to discover the states that can be visited during the plan execution requires foreknowing the conditions to be preserved for each environment state. We present here a new approach to discover the action policy for an environment under propositional constraints on states in MDP problems. The constraints are used by a strong probabilistic planning algorithm to reduce a state space whose transition probabilities are estimated by an action-learning reinforcement learning algorithm, thus simplifying the agents state space exploration and helping in the definition of the planning problem. The execution constraints, or preservation goals, comprised within the representation of the final goal, composes the extended reach ability goals. Experiments to validate the proposal were performed on an antenna coverage problem and produced interesting and promising results, demonstrating fast convergence to condition-preserving near-optimal policies that keep valid a set of propositions while reaching a final goal.

An Attentional Model for Autonomous Mobile Robots

The increase of applications that use autonomous robots has endowed them with a high number of sensors and actuators that are sometimes redundant. This new highly complex systems and the type of environment where they are expected to operate require them to deal with data overload and data fusion. In humans that face the same problem when sounds, images, and smells are presented to their sensors in a daily scene, a natural filter is applied: attention. Although there are many computational models that apply attentive systems to robotics, they usually are restricted to two classes of systems: 1)xa0those that have complex biologically based attentional visual systems and 2)xa0those that have simpler attentional mechanisms with a larger variety of sensors. This work proposes an attentional model inspired from biological systems and that supports a variety of robotics sensors. Furthermore, it discusses the possibility of using multiple sensors to define multiple features, with feature extraction modules that can handle exogenous and endogenous attentional processes. The experiments were performed in a simulated high-fidelity environment, and results have shown that the model proposed can account for detecting salient (bottom-up) and desired (top-down) stimuli according to the modeled features.

Sharing information on extended reachability goals over propositionally constrained multi-agent state spaces

By exchanging propositional information, agents can implicitly reduce large domain state spaces, a feature that is particularly attractive for Reinforcement Learning approaches. This paper proposes a learning technique that combines a Reinforcement Learning algorithm and a planner for propositionally constrained state spaces, that autonomously help agents to implicitly reduce the state space towards possible plans that lead to a goal whilst avoiding irrelevant or inadequate states. State space constraints are communicated among the agents using a common constraint set based on extended reachability goals. A performance evaluation against standard Reinforcement Learning techniques showed that by extending autonomous learning with propositional constraints updated along the learning process can produce faster convergence to optimal policies due to early state space reduction caused by shared information on state space constraints.

Sparse Sampling Action Values Initialized by a Compact Representation Technique

Most of the techniques proposed for problems involving mobile robots are specified in terms of optimal control of Markov decision processes (MDPs). However, the state space dimension explosion makes such tabular MDP-based solutions unfeasible. As an alternative to this, a planning technique based on sparse sampling (SSA) of simulated instances of a MDP model has been suggested. Because the execution time of this algorithm is exponential on the level of an exploration tree and on the number of samplings to be generated, this paper proposes a technique where leaves null-values in the SSA algorithm are substitute by meaningful values, acquired from any of the following approaches: 1) a simple environment reward distribution; 2) a standard reinforcement learning algorithm, and 3) a compact representation on a coarse state discretization for generating initial estimates of the action values. The experiments carried out showed that such information-based variants of SSA lead quickly to better results than the original technique.