A5023071960- Child and Animal Learning Development
- Cognitive Science and Mapping
- Neural Networks and Applications
- Language and cultural evolution
- Cognitive Science and Education Research
- Neuroscience and Music Perception
- Language Development and Disorders
- Diabetes Management and Research
- Neural dynamics and brain function
- Reading and Literacy Development
- Neurobiology of Language and Bilingualism
- Memory and Neural Mechanisms
- Cognitive and developmental aspects of mathematical skills
- Advanced Text Analysis Techniques
- Computability, Logic, AI Algorithms
- Domain Adaptation and Few-Shot Learning
- Diabetes and associated disorders
- Health disparities and outcomes
- Music and Audio Processing
- Categorization, perception, and language
- Evolutionary Game Theory and Cooperation
- Retirement, Disability, and Employment
- Face Recognition and Perception
- Motor Control and Adaptation
- Flood Risk Assessment and Management
Cardiff University
2018-2025
Wales Institute of Social and Economic Research, Data and Methods
2025
Université de Bourgogne
2013-2024
IRCCS Istituto Auxologico Italiano
2023
Centre National de la Recherche Scientifique
2008-2022
University of Bristol
2014-2022
University of Liège
1997-2019
Laboratoire d'étude de l'apprentissage et du développement
1982-2019
University Hospital of Wales
2018-2019
Cardiff and Vale University Health Board
2019
From the Publisher: Winner of Pulitzer Prize, this book applies Godel's seminal contribution to modern mathematics study human mind and development artificial intelligence.
Abstract High-level perception—the process of making sense complex data at an abstract, conceptual level—is fundamental to human cognition. Through high-level perception, chaotic environmental stimuli are organized into mental representations that used throughout cognitive processing. Much work in traditional artificial intelligence has ignored the by starting with hand-coded representations. In this paper, we argue dismissal perceptual processes leads distorted models We examine some...
Being able to automatically perceive a variety of emotions from text alone has potentially important applications in CMC and HCI that range identifying mood online posts enabling dynamically adaptive interfaces. However, such ability not been proven human raters or computational systems. Here we examine the naive emotion detect one eight emotional categories 50 200 word samples real blog text. Using expert as 'gold standard', naive-expert rater agreement increased with longer texts, was high...
Journal Article Subcognition and the Limits of TuringTest Get access ROBERT M. FRENCH Centre for Research on Concepts Cognition, Indiana UniversityBloomington, IN 47405, USA Search other works by this author on: Oxford Academic Google Scholar Mind, Volume XCIX, Issue 393, January 1990, Pages 53–65, https://doi.org/10.1093/mind/XCIX.393.53 Published: 01 1990
A differentially methylated region (DMR) is a genomic in which DNA methylation consistently positively or negatively associated with phenotype exposure. We demonstrate that existing algorithms for identifying DMRs either fail to control false positive rates (comb-p and DMRcate), suffer from low power (bumphunter) lack modeling flexibility (seqlm). introduce new method, dmrff, overcomes these shortcomings can additionally be used meta-analyze multiple datasets. When applied investigate...
Young infants show unexplained asymmetries in the exclusivity of categories formed on basis visually presented stimuli. A connectionist model is described that shows similar when categorizing same stimuli to infants. The can be explained terms an associative learning mechanism, distributed internal representations, and statistics feature distributions was used explore robustness this asymmetry. predicts asymmetry will persist a category acquired presence mixed exemplars. An experiment with...
Disentangling bottom-up and top-down processing in adult category learning is notoriously difficult. Studying infancy provides a simple way of exploring while minimizing the contribution information. Three- to 4-month-old infants presented with cat or dog images will form perceptual representation for that excludes dogs includes cats. The authors argue an inclusion relationship distribution features explains asymmetry. Using computational modeling behavioral testing, show asymmetry can be...
Abstract A major problem with connectionist networks is that newly-learned information may completely destroy previously-learned unless the network continually retrained on old information. This phenomenon, known as catastrophic forgetting, unacceptable both for practical purposes and a model of mind. paper advances claim forgetting in part result overlap system's distributed representations can be reduced by reducing this overlap. simple algorithm, called activation sharpening, presented...
Individuals of all ages extract structure from the sequences patterns they encounter in their environment, an ability that is at very heart cognition. Exactly what underlies this has been subject much debate over years. A novel mechanism, implicit chunk recognition (ICR), proposed for sequence segmentation and extraction. The mechanism relies on previously encountered subsequences (chunks) input rather than prediction upcoming items sequence. connectionist autoassociator model ICR, truncated...
Applications of multilevel models to continuous outcomes nearly always assume constant residual variance and random effects variances covariances. However, modeling heterogeneity can prove a useful indicator model misspecification, in some educational behavioral studies, it may even be direct substantive interest. The purpose this article is review, describe, illustrate set recent extensions two-level that allow the variance–covariance components specified as functions predictors. These...
The hostile environment that older lesbian, gay, bisexual and transgender (LGBT) people faced at younger ages in the United Kingdom (UK) may have a lasting negative impact on their health. This systematic scoping review adds to current knowledge base through comprehensively synthesising evidence what is known about extent nature of health care inequalities, as well highlighting gaps which point way towards future research priorities. We searched four databases, undertook manual searching,...
In order to solve the 'sensitivity-stability' problem-and its immediate correlate, problem of sequential learning-it is crucial develop connectionist architectures that are simultaneously sensitive to, but not excessively disrupted by, new input. French (1992) suggested alleviate a particularly severe form this disruption, catastrophic forgetting, it was necessary for networks separate dynamically their internal representations during learning. McClelland et al. (1995) went even further....
Emotion is central to human interactions, and automatic detection could enhance our experience with technologies. We investigate the linguistic expression of fine-grained emotion in 50 200 word samples real blog texts previously coded by expert naive raters. Content analysis (LIWC) reveals angry authors use more affective language negative affect words, that joyful positive words. Additionally, a co-occurrence semantic space approach (LSA) was able identify fear (which raters not do). relate...