1,721,066 research outputs found
Object-centred neglect: Simulation with head-centred coding based on Gaussian gaze-dependent units
No full textRight brain damaged patients affected by contralesional object-centred neglect are able to process all objects around them but systematically omit the left part of these objects. We show that pure object-centred neglect can be simulated by a basis function neural network in which the activity of units allowing head-centred coding of space is based on the activity of gaze-dependent units with no lateral gradient of preferred eye positions along the horizontal space. This type of network simulates complete dissociation between object-centred and egocentric neglect, as observed in patients' copies of multi-item drawings. Based on these results and available neurophysiological, clinical and brain imaging data, it is proposed that disruption of a cerebral network including dorsal occipital and parietal areas and the supplementary eye fields could be the main cause of object-centred neglect. (c) 2007 Elsevier Ltd. All rights reserved
Manuale di Scienza Cognitiva: Intelligenza Artificiale classica e Psicologia Cognitiva
No full textIl libro descrive gli sviluppi fondamentali della Scienza Cognitiva, sia sul piano filosofico, che della Psicologia Cognitiva e dell'Intelligenza Artificiale tradizionale. Sono descritti in dettaglio modelli formalizzati di apprendimento, memoria, attenzione, percezione, formazione e uso dei concetti, soluzione dei problem
“Complexity and Paternalism”
No full textThe aim of the paper is to discuss paternalistic economic policy in relation to complexity. It attempts to do so by examining how complexity affects economic inquiry as a whole. The point of departure is a number of papers written by David Colander, where he contends that progress in mathematics and computational technology provides a new outlook on economic policy, making it reasonable to advocate “libertarian paternalism”.
Colander focuses on complexity in economic dynamics. I argue, however, that he neglects a range of issues associated to complexity in cognition. My contention is that both individuals and policy makers do not only need to make choices; they also need to choose how to choose. The discretionary nature of these choices reasserts the value-laden nature not only of economic policy but of the economic inquiry that underlies it. When these issues are taken into account, Colander’s view of public policy and paternalism turns out to be too simple and a different notion of paternalism is required.
The paper is structured as follows. Following a brief outline of Colander’s views I discuss a few features of cognitive complexity in order to point to the shortcomings of libertarian paternalism. I then introduce the notion of a knowledge context and argue that it can be affected by the purposive action of vested interests. Finally I outline the characteristics of two possible forms of paternalistic policy
Natural Rates of Teachers' Approval and Disapproval in Italian Primary and Secondary Schools Classroom
No full textSince the 1960s, researchers have been demonstrating the power of teacher approval on the behaviour of students. Despite the huge amount of research in English-speaking countries the natural rates of approval and disapproval had never been investigated in Italian schools. The aims of this study were: (a) to examine the proportionality of different types of verbal feedback used by teachers; (b) to examine the relationship between teachers use of verbal feedback and the behaviour of students. A total of 314 observations in schools were conducted. As regards teacher verbal behaviour: we found that both approval and disapproval rates per minute declined and neutral verbal behaviour (instructions, expositions, explanations) increased as the age of the pupils increases. There was a clear difference between primary school and secondary school teachers
Decomposing Dynamical Systems
No full textDynamical systems on monoids have been recently proposed as minimal mathematical models for the intuitive notion of deterministic dynamics. This paper shows that any dynamical system DS_L on a monoid L can be exhaustively decomposed into a family of mutually disconnected subsystems—the constituent systems of DS_L . In addition, constituent systems are themselves indecomposable, even though they may very well be complex. Finally, this work also makes clear how any dynamical system DS_L turns out to be identical to the sum of all its constituent systems. Constituent systems can thus be thought as the indecomposable, but possibly complex, building blocks to which the dynamics of an arbitrary complex system fully reduces. However, no further reduction of the constituents is possible, even if they are themselves complex
Personality and complex systems. An expanded view
No full textNowadays, dynamic S-P-R models, developed within personality research, seem to offer the chance for the development of a unitary research framework, one including individual and environmental factors, from both structural and dynamic levels of personality. Complexity concept is a fundamental element of convergence of the different existing theoretical and methodological approaches. From this expanded view, personality appears as an “hyper-complex” system of intra-individual circular interactions, and of recursive relations between internal and external factors, that develops and grows self-organizing. This approach takes into account, within a wide interpretative framework, the contributions from various humanistic and scientific disciplines
Are dynamically undecidable systems ubiquitous?
No full textStephen Wolfram has maintained that almost any system whose behavior is not obviously simple is computationally universal and, consequently, its long term behavior is undecidable. Wolfram's tenet is a direct consequence of his Principle of Computational Equivalence (PCE). In this paper, I propose an independent argument for the ubiquity of computational universality and, as a consequence, dynamical undecidability as well. My argument does not presuppose PCE and, in essence, it is based on the recognition of two facts: (1) the existence of a strong structural similarity between the transition graphs of any two computational systems; (2) the mapping needed for computational universality is emulation, which is itself a quite weak structural mapping
- …
