Certain rough set lattices induced by binary relations form almost disjoint unions of Boolean algebras, a subclass of orthomodular lattices that captures quantum-like structure. In empirical data analysis, however, rough set lattices rarely realize this ideal form exactly. To quantify approximate similarity to such structures, we introduce numerical indices for rough set lattices. Simulation experiments based on controlled perturbations of binary relations show that the proposed indices distinguish structured perturbations from density-matched random baselines and remain informative under substantial perturbation. We further apply the framework to rough set lattices derived from machine-learning confusion matrices. In the examples studied, ordinary classification outcomes do not exhibit the characteristic signatures of almost disjoint unions of Boolean algebras, whereas a superposition-based setting does. The proposed approach provides a quantitative methodology for detecting latent lattice-theoretic organization in complex and intelligent systems, with potential relevance to cognitive and other self-organizing information systems.

Reservoir computing (RC) is a computer architecture allowing us to utilize any nonlinear dynamical system as a computing device called a reservoir. Exploiting elementary cellular automata (ECAs) as reservoirs is no exception. Recent studies have shown that the RC implemented by asynchronously tuned ECAs (AT_ECAs) has an advantage of enhancing the learning ability to identify multiple patterns, suggesting effects of critical spacetime patterns that the AT_ECAs create in a wide range of transition rules. However, the generalization capability of the AT_ECA-based RC remains unclear. This study evaluated the generalization performance of the AT_ECA-based RC using the temporal parity task in comparison with the ECA-based RC. We found that the AT_ECA-based RC demonstrated higher performance than ECA-based RC in most rules. This might have been a result of critical behaviors that the AT_ECAs universally generate with different mechanisms from the ECAs.

For a reinforcement learning agent to finish trial-and-error in a realistic time duration, it is necessary to limit the scope of exploration during the learning process. However, limiting the exploration scope means limitation in optimality: the agent could fall into a suboptimal solution. This is the nature of local, bottom-up way of learning. An alternative way is to set a goal to be achieved, which is a more global, top-down way. The risk-sensitive satisficing (RS) value function incorporate, as a method of the latter way, the satisficing principle into reinforcement learning and enables agents to quickly converge to exploiting the optimal solution without falling into a suboptimal one, when an appropriate goal (aspiration level) is given. However, how best to determine the aspiration level is still an open problem. This study proposes social satisficing, a framework for multi-agent reinforcement learning which determines the aspiration level through information sharing among multiple agents. In order to verify the effectiveness of this novel method, we conducted simulations in a learning environment with many suboptimal goals (SuboptimaWorld). The results show that the proposed method, which converts the aspiration level at the episodic level into local (state-wise) aspiration levels, possesses a higher learning efficiency than any of the compared methods, and that the novel method has the ability to autonomously adjust exploration scope, while keeping the shared information minimal. This study provides a glimpse into an aspect of human and biological sociality which has been mentioned little in the context of artificial intelligence and machine learning.

Elementary cellular automata (ECAs) generate critical spacetime patterns in a few local rules, which are expected to have advantages in reservoir computing (RC). However, previous studies have not revealed the advantages of critical spacetime patterns in RC. In this paper, we focus on the distractor’s length in the time series data for learning and clarify the advantages of the critical spacetime patterns. Furthermore, we propose asynchronously tuned ECAs (AT_ECAs) to generate universally critical spacetime patterns in many local rules. Based on the results achieved in this study, we propose RC based on AT_ECAs. Moreover, we show that the universal criticality of AT_ECAs is effective for learning time series data.

Asynchronously tuned elementary cellular automata (AT-ECA) are described with respect to the relationship between active and passive updating, and that spells out the relationship between synchronous and asynchronous updating. Mutual tuning between synchronous and asynchronous updating can be interpreted as the model for dissipative structure, and that can reveal the critical property in the phase transition from order to chaos. Since asynchronous tuning easily makes behavior at the edge of chaos, the property of AT-ECA is called the unfolded edge of chaos. The computational power of AT-ECA is evaluated by the quantitative measure of computational universality and efficiency. It shows that the computational efficiency of AT-ECA is much higher than that of synchronous ECA and asynchronous ECA.

Although natural and bioinspired computing has developed significantly, the relationship between the computational universality and efficiency beyond the Turing machine has not been studied in detail. Here, we investigate how asynchronous updating can contribute to the universal and efficient computation in cellular automata (CA). First, we define the computational universality and efficiency in CA and show that there is a trade-off relation between the universality and efficiency in CA implemented in synchronous updating. Second, we introduce asynchronous updating in CA and show that asynchronous updating can break the trade-off found in synchronous updating. Our finding spells out the significance of asynchronous updating or the timing of computation in robust and efficient computation.

The asynchronously tuned cellular automaton (AT_ECA) we proposed has been shown to generate critical spatiotemporal patterns without fine-tuning of order parameters. In this study, we propose a learning system that applies the remarkable characteristics of AT_ECA to reservoir computing, which has recently been attracting attention as a learning model for time series data. Then, the learning ability of the proposed system was evaluated by the learning task called five-bit task. As a result, it became clear that the success rate of learning is relatively high even if distractor in time series data to learn is long.

In this paper, we propose asynchronously tuned elementary cellular automata (AT_ECA) as models that implement a new type of self-organized criticality (SOC). SOC in AT_ECA is based on asynchronously updating and locally tuning the consistency between dual modes of transition. A previous work showed that AT_ECA generate class 4-like spacetime patterns over a wide area of the rule space, and the density decay follows a power law for some of the rules. In this study, we performed a spectral analysis of AT_ECA, of which a great number of rules were found to exhibit 1/f noise, suggesting that AT_ECA realize critical states without selecting specific rules or fine-tuning parameters.

We propose a method of applying lattice algebra to the analysis of multivariate time series data. We measured the body motion of a Shorinji-Kempo kata using acceleration sensors at five positions on the body. The proposed analysis was applied to the time series data obtained from the sensors. As a result, a correlation was observed between the skill levels and the number of elements of the lattice generated using the time series data. We observed that highly skilled subjects executed more complex motions; thus, we consider the number of lattice elements as an index of complexity of the space-time pattern.

As the scope of reinforcement learning broadens, the number of possible states and of executable actions, and hence the product of the two sets explode. Often, there are more feasible options than allowed trials, because of physical and computational constraints imposed on the agents. In such an occasion, optimization procedures that require first trying all the options once do not work. The situation is what the theory of bounded rationality was proposed to deal with. We formalize the central heuristics of bounded rationality theory named satisficing. Instead of the traditional formulation of satisficing at the policy level in terms of reinforcement learning, we introduce a value function that implements the asymmetric risk attitudes characteristic of human cognition. Operated under the simple greedy policy, the RS (reference satisficing) value function enables an efficient satisficing in K-armed bandit problems, and when the reference level for satisficing is set at an appropriate value, it leads to effective optimization. RS is also tested in a robotic motion learning task in which a robot learns to perform giant-swings (acrobot). While the standard algorithms fail because of the coarse-grained state space, RS shows a stable performance and autonomous exploration that goes without randomized exploration and its gradual annealing necessary for the standard methods.

Some of the authors have previously proposed a cognitively inspired reinforcement learning architecture (LS-Q) that mimics cognitive biases in humans. LS-Q adaptively learns under uniform, coarse-grained state division and performs well without parameter tuning in a giant-swing robot task. However, these results were shown only in simulations. In this study, we test the validity of the LS-Q implemented in a robot in a real environment. In addition, we analyze the learning process to elucidate the mechanism by which the LS-Q adaptively learns under the partially observable environment. We argue that the LS-Q may be a versatile reinforcement learning architecture, which is, despite its simplicity, easily applicable and does not require well-prepared settings.

The arch structure of human foot plays important roles such as impact absorption for bipedal locomotion. In this paper, the two-arch structure is introduced into the foot model of biped robot. Using a flat foot model, a one-arch foot model and the two-arch foot model, simulations of robot motion are performed by ODE and effects of the arch structure are verified. When the inner arch was softer than the outer arch like human, the robot motion such that the COG returns to inner direction was observed. It was verified that the property of the returning motion was determined by the difference of elasticities between the inner arch and the outer arch, and the property of motion in the frontal direction was determined by the average of elasticities.

In this paper, aiming at human-robot cooperation using electroencephalogram (EEG), we studied on EEG when the load changed during a task. EEG and electromyogram (EMG) signals were recorded during a crank rotation task, where the load of the task was changed at random timing. Frequency spectra of the recorded EEG signals were analyzed, and difference between ON and OFF of the load was observed.

Sphere-shaped robots are expected to be useful, in that they can move around in home or office environments without getting stuck, even in a dusty room with a lot of obstacles. In order to provides both omni-directional mobility and a large inner space, the authors have previously proposed a new driving scheme employing an omni-directional three-wheel internal unit. Then, a prototype robot was fabricated and constant speed movements were achieved experimentally. However, due to the phere shape, it is observed that the robot shakes largely when it stops running. Therefore in this presentation, reduction of swing movements after stopping is considered. First, a scheme is proposed in which the rolling and pitching angular velocities of the internal unit are measured using a gyro sensor and feedback to velocity reference signals for omni-wheels. Second, the scheme is implemented on the prototype robot. Finally, Experiments using the prototype are described which confirms the validity of the proposed scheme.

Many algorithms and methods in artificial intelligence or machine learning were inspired by human cognition. As a mechanism to handle the exploration–exploitation dilemma in reinforcement learning, the loosely symmetric (LS) value function that models causal intuition of humans was proposed (Shinohara et al., 2007). While LS shows the highest correlation with causal induction by humans, it has been reported that it effectively works in multi-armed bandit problems that form the simplest class of tasks representing the dilemma. However, the scope of application of LS was limited to the reinforcement learning problems that have K actions with only one state (K-armed bandit problems). This study proposes LS-Q learning architecture that can deal with general reinforcement learning tasks with multiple states and delayed reward. We tested the learning performance of the new architecture in giant-swing robot motion learning, where uncertainty and unknown-ness of the environment is huge. In the test, the help of ready-made internal models or functional approximation of the state space were not given. The simulations showed that while the ordinary Q-learning agent does not reach giant-swing motion because of stagnant loops (local optima with low rewards), LS-Q escapes such loops and acquires giant-swing. It is confirmed that the smaller number of states is, in other words, the more coarse-grained the division of states and the more incomplete the state observation is, the better LS-Q performs in comparison with Q-learning. We also showed that the high performance of LS-Q depends comparatively little on parameter tuning and learning time. This suggests that the proposed method inspired by human cognition works adaptively in real environments.

Distributed connectionist networks have difficulty learning incrementally because the representations in the network overlap. Therefore, it is necessary to reduce the overlaps of representations for incremental learning. At the same time, the representational overlaps give these networks the ability to generalize. In this study, we use a modified multilayered neural network to numerically examine the trade-off between incremental learning and generalization abilities, and then we propose a novel network model with structural lateral inhibitions to reconcile the two abilities. We also analyze the behavior of the proposed model using Formal Concept Analysis, which reveals that the network implements "conceptualization": differentiation and meditation between intensional and extensional representations. This study suggests a new paradigm for the traditional question, whether representations in the brain are distributed or not.

Sphere-shaped robots are expected to be useful in that they can move around in home or office environments without getting stuck even in a dusty room with a lot of obstacles. Besides, it is desirable that one can move omni-directionally and that its body has enough space to carry additional functions such as sensors. Until now, various types of sphere-shaped robot have been proposed and studied. However, few of them satisfy both the requirements at the same time. Thus, this study proposes a new driving scheme employing an omni-directional three-wheel internal unit, in order to make it capable of moving omni-directionally providing large inner space. Then, for constant velocity movements, conditions for speed references of three wheels are derived, so that the internal unit runs smoothly without making any slippage against the inner surface of the sphere. Finally, a prototype robot is developed and the validity of the conditions is examined through experiments.

Symmetric cognitive biases are human inferences that induce "q given p" from condition "p given q". LS model represents these human tendencies in human-like decision-making, which had been proven a great performance in the n-armed bandit problem. In the previous study, we proved it effective in robot motion learning as well by simulation. In this study, we verified the effectiveness by using a real giant-swing robot which we made.