Recently, a lot of researches on Brain-Computer Interfaces (BCIs) have been reported. It is expected that BCIs will help patients like those with Amyotrophic Lateral Sclerosis (ALS) to control a wheel chair or to communicate with other people just by using thoughts. BCIs analyze biological signals such as Electroencephalogram (EEG) to recognize thoughts, and especially those associated with motor imagery have been well investigated. The Bereitschschafts Potential (BP) is one of the features of EEGs related to body movements, which occurs before voluntary limb movements. The BPs associated with different movements have different temporal patterns. This paper proposes a method to capture the difference of the BP patterns. In this paper, the proposed method is applied to EEGs of motor imagery used in BCI Competition III and the effectivenesss of the proposed method is shown.
In this paper, a model of fuzzy category formation was presented. This model achieves fuzzy clustering of exemplars and reduction of feature dimensions simultaneously. Human category formation is a unsupervised learning problem. Forming categories is computing optimal partitions from similarity between exemplars. Category formation is thought to be a clustering humans do. Presented model has two assumptions to characterize the model as human category formation. First, humans reduce feature dimensions and choose informative features to form category efficiently. When we observe exemplars, large amount of features are extracted. We can obtain a good result of category formation, if we used all features. However, because of the limited cognitive resources, it is implausible that humans use large amount of features for category formation. Second, human category structure has fuzziness. In many cases, there are no necessary and sufficient defining features to characterize a category. Features shared by many exemplars form categories. From this family resemblance point of view, human category can't stand up without fuzziness. In this paper, reduced k-means method, which achieves crisp clustering and dimension reduction simultaneously, was extended to represent fuzzy category by using entropy regularization method. Then presented model's validity and limitations as a cognitive model are discussed from results of two simulations and rating data analysis.
We conducted a psychological study to examine vague judgments in social and perceptual tasks. Two hundred and fifty-two subjects were randomly divided into three groups of roughly equal size. Group 1 was told to make an accurate estimate while group 2 was told to make a vague/rough estimate and group 3 was told to make an estimate with a fuzzy rating method. The results indicated that participants told to estimate with accuracy were better than those with vague condition for two of the perceptual tasks (square area and length of vertical line) but were not significantly different for three of the perceptual tasks (number of dots and length of horizontal line). We also found that participants in the vague condition were more accurate than the accuracy condition for one social task (number of suicides per year), but not significantly different for 2 of the social tasks (number of marriages per year and height of Mt. Fuji). The results suggested that vague judgment was accurate and adaptive strategy in some situations.
Communication is a basis of social ability in human. It is an exchange of information among multiple entities, and within this, ambiguity has been thought to be unnecessary for communication. This paper proposed a hypothesis that ambiguity is included in interpersonal communication and that it plays a critical role, not only in one-to-one relationship but also in a large scale social group. In order to test this hypothesis, we conducted one simulation study and two behavioral experiments. In the paper we showed the detail evidences supporting the hypothesis and discussed about a neural mechanism for disambiguating process in communication ability.
This study analyzed modes of reasoning on the Bayesian problem which could be solved with a basic first-order quantification of probability. Junior high school students (n=33) and university students (n=48) participated in this study. The results were as follows: The Bayesian problem which could be solved with a basic first-order quantification of probability (type 1 problem) was more difficult than the basic first-order quantification of probability problem (type 2 problem) but easier than the original Bayesian problem (type 3 problem). It demonstrates that the level of difficulty of problems depends on how they are framed. In addition, although the same modes of reasoning appeared, the cognitive processes behind them were different between junior high school students and university students. It suggests the importance of a developmental perspective in researches on judgment under ambiguity. Ambiguity can be interpreted as a perturbation factor that has effects on the function of the cognitive system (Piaget, 1970/2007) in the Ellsberg paradox (Ellsberg, 1961), while the framing can be interpreted as such a factor in the problems used in this study. In other words, this study demonstrates that judgments deviate from the normative theory of probability under unambiguity as judgments deviate from the independence axiom of expected utility theory under ambiguity.
Matrix operations derived from t-norm and t-conorm are proposed. The algebraic properties of the matrix operations are studied. Using the matrix as an adjacent matrix of a fuzzy graph, a numerical trust evaluation of networks is performed. The fuzziness of trustability distribution is calculated.
A Choquet integral model with multiple outputs is proposed. For an input pair (x1,..., xn) and m fuzzy measures μj, the output values y1,..., ym are calculated by the extended Choquet integral with respect to the j-th fuzzy measure μj. By imposing the constrains that the sum of μ(A)=1 for all A, the sum of the output values, ∑yj is founded to be 1. By using this model, it is possible to categorize objects with fuzzy inputs into m classes. We show an example where objects are categorized as class1, class2, contradiction class, and rest-unknown class.