EEG decoding of semantic category reveals distributed representations for single concepts

Achieving a clearer picture of categorial distinctions in the brain is essential for our understanding of the conceptual lexicon, but much more fine-grained investigations are required in order for this evidence to contribute to lexical research. Here we present a collection of advanced data-mining techniques that allows the category of individual concepts to be decoded from single trials of EEG data. Neural activity was recorded while participants silently named images of mammals and tools, and category could be detected in single trials with an accuracy well above chance, both when considering data from single participants, and when group-training across participants. By aggregating across all trials, single concepts could be correctly assigned to their category with an accuracy of 98%. The pattern of classifications made by the algorithm confirmed that the neural patterns identified are due to conceptual category, and not any of a series of processing-related confounds. The time intervals, frequency bands and scalp locations that proved most informative for prediction permit physiological interpretation: the widespread activation shortly after appearance of the stimulus (from 100. ms) is consistent both with accounts of multi-pass processing, and distributed representations of categories. These methods provide an alternative to fMRI for fine-grained, large-scale investigations of the conceptual lexicon. © 2010 Elsevier Inc.

Spatial attention enhances object coding in local and distributed representations of the lateral occipital complex

The modulation of neural activity in visual cortex is thought to be a key mechanism of visual attention. The investigation of attentional modulation in high-level visual areas, however, is hampered by the lack of clear tuning or contrast response functions. In the present functional magnetic resonance imaging study we therefore systematically assessed how small voxel-wise biases in object preference across hundreds of voxels in the lateral occipital complex were affected when attention was directed to objects. We found that the strength of attentional modulation depended on a voxel's object preference in the absence of attention, a pattern indicative of an amplificatory mechanism. Our results show that such attentional modulation effectively increased the mutual information between voxel responses and object identity. Further, these local modulatory effects led to improved information-based object readout at the level of multi-voxel activation patterns and to an increased reproducibility of these patterns across repeated presentations. We conclude that attentional modulation enhances object coding in local and distributed object representations of the lateral occipital complex.

Linear Classifiers Based on Binary Distributed Representations

Binary distributed representations of vector data (numerical, textual, visual) are investigated in classification tasks. A comparative analysis of results for various methods and tasks using artificial and real-world data is given.

Learning user and product distributed representations using a sequence model for sentiment analysis

In product reviews, it is observed that the distribution of polarity ratings over reviews written by different users or evaluated based on different products are often skewed in the real world. As such, incorporating user and product information would be helpful for the task of sentiment classification of reviews. However, existing approaches ignored the temporal nature of reviews posted by the same user or evaluated on the same product. We argue that the temporal relations of reviews might be potentially useful for learning user and product embedding and thus propose employing a sequence model to embed these temporal relations into user and product representations so as to improve the performance of document-level sentiment analysis. Specifically, we first learn a distributed representation of each review by a one-dimensional convolutional neural network. Then, taking these representations as pretrained vectors, we use a recurrent neural network with gated recurrent units to learn distributed representations of users and products. Finally, we feed the user, product and review representations into a machine learning classifier for sentiment classification. Our approach has been evaluated on three large-scale review datasets from the IMDB and Yelp. Experimental results show that: (1) sequence modeling for the purposes of distributed user and product representation learning can improve the performance of document-level sentiment classification; (2) the proposed approach achieves state-of-The-Art results on these benchmark datasets.

Improving distributed representation of word sense via WordNet Gloss composition and context clustering

In recent years, there has been an increas-ing interest in learning a distributed rep-resentation of word sense. Traditional context clustering based models usually require careful tuning of model parame-ters, and typically perform worse on infre-quent word senses. This paper presents a novel approach which addresses these lim-itations by first initializing the word sense embeddings through learning sentence-level embeddings from WordNet glosses using a convolutional neural networks. The initialized word sense embeddings are used by a context clustering based model to generate the distributed representations of word senses. Our learned represen-tations outperform the publicly available embeddings on 2 out of 4 metrics in the word similarity task, and 6 out of 13 sub tasks in the analogical reasoning task.

A gloss composition and context clustering based distributed word sense representation model

In recent years, there has been an increasing interest in learning a distributed representation of word sense. Traditional context clustering based models usually require careful tuning of model parameters, and typically perform worse on infrequent word senses. This paper presents a novel approach which addresses these limitations by first initializing the word sense embeddings through learning sentence-level embeddings from WordNet glosses using a convolutional neural networks. The initialized word sense embeddings are used by a context clustering based model to generate the distributed representations of word senses. Our learned representations outperform the publicly available embeddings on half of the metrics in the word similarity task, 6 out of 13 sub tasks in the analogical reasoning task, and gives the best overall accuracy in the word sense effect classification task, which shows the effectiveness of our proposed distributed distribution learning model.

Eigen-Adaptation and Distributed Representation of 2-D Phase, Energy, Scale and Orientation in Spatial Vision

Distributed representations (DR) of cortical channels are pervasive in models of spatio-temporal vision. A central idea that underpins current innovations of DR stems from the extension of 1-D phase into 2-D images. Neurophysiological evidence, however, provides tenuous support for a quadrature representation in the visual cortex, since even phase visual units are associated with broader orientation tuning than odd phase visual units (J.Neurophys.,88,455–463, 2002). We demonstrate that the application of the steering theorems to a 2-D definition of phase afforded by the Riesz Transform (IEEE Trans. Sig. Proc., 49, 3136–3144), to include a Scale Transform, allows one to smoothly interpolate across 2-D phase and pass from circularly symmetric to orientation tuned visual units, and from more narrowly tuned odd symmetric units to even ones. Steering across 2-D phase and scale can be orthogonalized via a linearizing transformation. Using the tiltafter effect as an example, we argue that effects of visual adaptation can be better explained by via an orthogonal rather than channel specific representation of visual units. This is because of the ability to explicitly account for isotropic and cross-orientation adaptation effect from the orthogonal representation from which both direct and indirect tilt after-effects can be explained.

Cognitive neuroimaging: Cognitive science out of the armchair

Cognitive scientists were not quick to embrace the functional neuroimaging technologies that emerged during the late 20th century. In this new century, cognitive scientists continue to question, not unreasonably, the relevance of functional neuroimaging investigations that fail to address questions of interest to cognitive science. However, some ultra-cognitive scientists assert that these experiments can never be of relevance to the study of cognition. Their reasoning reflects an adherence to a functionalist philosophy that arbitrarily and purposefully distinguishes mental information-processing systems from brain or brain-like operations. This article addresses whether data from properly conducted functional neuroimaging studies can inform and subsequently constrain the assumptions of theoretical cognitive models. The article commences with a focus upon the functionalist philosophy espoused by the ultra-cognitive scientists, contrasting it with the materialist philosophy that motivates both cognitive neuroimaging investigations and connectionist modelling of cognitive systems. Connectionism and cognitive neuroimaging share many features, including an emphasis on unified cognitive and neural models of systems that combine localist and distributed representations. The utility of designing cognitive neuroimaging studies to test (primarily) connectionist models of cognitive phenomena is illustrated using data from functional magnetic resonance imaging (fMRI) investigations of language production and episodic memory.

The Physiology and Computation of Pyramidal Neurons

A variety of neural signals have been measured as correlates to consciousness. In particular, late current sinks in layer 1, distributed activity across the cortex, and feedback processing have all been implicated. What are the physiological underpinnings of these signals? What computational role do they play in the brain? Why do they correlate to consciousness? This thesis begins to answer these questions by focusing on the pyramidal neuron. As the primary communicator of long-range feedforward and feedback signals in the cortex, the pyramidal neuron is set up to play an important role in establishing distributed representations. Additionally, the dendritic extent, reaching layer 1, is well situated to receive feedback inputs and contribute to current sinks in the upper layers. An investigation of pyramidal neuron physiology is therefore necessary to understand how the brain creates, and potentially uses, the neural correlates of consciousness. An important part of this thesis will be in establishing the computational role that dendritic physiology plays. In order to do this, a combined experimental and modeling approach is used.

This thesis beings with single-cell experiments in layer 5 and layer 2/3 pyramidal neurons. In both cases, dendritic nonlinearities are characterized and found to be integral regulators of neural output. Particular attention is paid to calcium spikes and NMDA spikes, which both exist in the apical dendrites, considerable distances from the spike initiation zone. These experiments are then used to create detailed multicompartmental models. These models are used to test hypothesis regarding spatial distribution of membrane channels, to quantify the effects of certain experimental manipulations, and to establish the computational properties of the single cell. We find that the pyramidal neuron physiology can carry out a coincidence detection mechanism. Further abstraction of these models reveals potential mechanisms for spike time control, frequency modulation, and tuning. Finally, a set of experiments are carried out to establish the effect of long-range feedback inputs onto the pyramidal neuron. A final discussion then explores a potential way in which the physiology of pyramidal neurons can establish distributed representations, and contribute to consciousness.

Understanding deep architectures and the effect of unsupervised pre-training

Cette thèse porte sur une classe d'algorithmes d'apprentissage appelés architectures profondes. Il existe des résultats qui indiquent que les représentations peu profondes et locales ne sont pas suffisantes pour la modélisation des fonctions comportant plusieurs facteurs de variation. Nous sommes particulièrement intéressés par ce genre de données car nous espérons qu'un agent intelligent sera en mesure d'apprendre à les modéliser automatiquement; l'hypothèse est que les architectures profondes sont mieux adaptées pour les modéliser. Les travaux de Hinton (2006) furent une véritable percée, car l'idée d'utiliser un algorithme d'apprentissage non-supervisé, les machines de Boltzmann restreintes, pour l'initialisation des poids d'un réseau de neurones supervisé a été cruciale pour entraîner l'architecture profonde la plus populaire, soit les réseaux de neurones artificiels avec des poids totalement connectés. Cette idée a été reprise et reproduite avec succès dans plusieurs contextes et avec une variété de modèles. Dans le cadre de cette thèse, nous considérons les architectures profondes comme des biais inductifs. Ces biais sont représentés non seulement par les modèles eux-mêmes, mais aussi par les méthodes d'entraînement qui sont souvent utilisés en conjonction avec ceux-ci. Nous désirons définir les raisons pour lesquelles cette classe de fonctions généralise bien, les situations auxquelles ces fonctions pourront être appliquées, ainsi que les descriptions qualitatives de telles fonctions. L'objectif de cette thèse est d'obtenir une meilleure compréhension du succès des architectures profondes. Dans le premier article, nous testons la concordance entre nos intuitions---que les réseaux profonds sont nécessaires pour mieux apprendre avec des données comportant plusieurs facteurs de variation---et les résultats empiriques. Le second article est une étude approfondie de la question: pourquoi l'apprentissage non-supervisé aide à mieux généraliser dans un réseau profond? Nous explorons et évaluons plusieurs hypothèses tentant d'élucider le fonctionnement de ces modèles. Finalement, le troisième article cherche à définir de façon qualitative les fonctions modélisées par un réseau profond. Ces visualisations facilitent l'interprétation des représentations et invariances modélisées par une architecture profonde.

Patterns of regional brain hypometabolism associated with knowledge of semantic features and categories in Alzheimer's disease

The study of semantic memory in patients with Alzheimer's disease (AD) has raised important questions about the representation of conceptual knowledge in the human brain. It is still unknown whether semantic memory impairments are caused by localized damage to specialized regions or by diffuse damage to distributed representations within nonspecialized brain areas. To our knowledge, there have been no direct correlations of neuroimaging of in vivo brain function in AD with performance on tasks differentially addressing visual and functional knowledge of living and nonliving concepts. We used a semantic verification task and resting 18-fluorodeoxyglucose positron emission tomography in a group of mild to moderate AD patients to investigate this issue. The four task conditions required semantic knowledge of (1) visual, (2) functional properties of living objects, and (3) visual or (4) functional properties of nonliving objects. Visual property verification of living objects was significantly correlated with left posterior fusiform gyrus metabolism (Brodmann's area [BA] 37/19). Effects of visual and functional property verification for non-living objects largely overlapped in the left anterior temporal (BA 38/20) and bilateral premotor areas (BA 6), with the visual condition extending more into left lateral precentral areas. There were no associations with functional property verification for living concepts. Our results provide strong support for anatomically separable representations of living and nonliving concepts, as well as visual feature knowledge of living objects, and against distributed accounts of semantic memory that view visual and functional features of living and nonliving objects as distributed across a common set of brain areas.

COGNITIVE IMPACT OF INTERACTIVE MULTIMEDIA

Technology has been gradually introduced in heath education. One of the most attractive features of this technology-based education is the use of multimedia. In this article we explore the research evidence about the role that multimedia is playing in education. From that analysis we describe the most relevant features of this technology to prepare a common ground of discussion about the evaluation of its impact on educational outcomes. As part of this analysis, we organize current research evidence on the use of technology in medical education, distinguishing diverse variables involved in the process, like knowledge (declarative, procedural), learner characteristics, curricular scenario, etc. This article presents an overview of the Distributed Representations theory and its relationship with research on educational outcomes and multimedia. Next we discuss the relationship between media and diverse learning theories, proposing a theory based taxonomy for educational multimedia.

Cognitive neuroimaging: Cognitive science out of the armchair

Cognitive scientists were not quick to embrace the functional neuroimaging technologies that emerged during the late 20th century. In this new century, cognitive scientists continue to question, not unreasonably, the relevance of functional neuroimaging investigations that fail to address questions of interest to cognitive science. However, some ultra-cognitive scientists assert that these experiments can never be of relevance to the Study of cognition. Their reasoning reflects an adherence to a functionalist philosophy that arbitrarily and purposefully distinguishes mental information-processing systems from brain or brain-like operations. This article addresses whether data from properly conducted functional neuroimaging studies can inform and Subsequently constrain the assumptions of theoretical cognitive models. The article commences with a focus upon the functionalist philosophy espoused by the ultra-cognitive scientists, contrasting it with the materialist philosophy that motivates both cognitive neuromiaging investigations and connectionist modelling of cognitive systems. Connectionism and cognitive neuroimaging share many features, including an emphasis on unified cognitive and neural models of systems that combine localist and distributed representations. The utility of designing cognitive neuroimaging studies to test (primarily) connectionist models of cognitive phenomena is illustrated using data from functional magnetic resonance imaging (fMRI) investigations of language production and episodic memory. (C) 2005 Elsevier Inc. All rights reserved.

Searching for Nearest Strings with Neural-Like String Embedding

We analyze an approach to a similarity preserving coding of symbol sequences based on neural distributed representations and show that it can be viewed as a metric embedding process.