Tag Info

Hot answers tagged

20

Probabilistic approaches of this sort are usually referred to more specifically as the bayesian approach and Chater and Tanenbaum are definitely bayesians (I have not read much by Yuille and can't comment). Bayesianism is more than just increasing in popularity and being encouraged; it is considered one of the big-4 approaches to cognitive-modeling, with the ...


14

When performing certain tasks, people’s inferences approximate Bayesian inference to a remarkable degree. For example, when people receive both haptic and visual information about the size of an object, they combine this information in a manner that very closely resembles Bayesian inference, taking account of the uncertainties associated with the visual and ...


14

Treisman & Gelade's Feature Integration Theory suggests that we are able to process an entire visual scene in parallel at the level of individual features. For example, in a visual search task, the time required to find a blue circle in a field of red circles is independent of the total number of circles. However, focused attention (typically foveal) is ...


13

Artem gave a very good answer, but I want to add one more weaknesses of probabilistic/Bayesian models: they are not mechanistic. This is related to Artem's point about neural grounding, but is a little different. The issue is that probabilistic models don't really provide insight into the underlying mechanism that produces the observed behavior -- if you ask ...


13

In my experience, the term "semantic knowledge" (or semantic memory or conceptual knowledge) is generally used to refer to knowledge of objects, word meanings, facts and people, without connection to any particular time or place. The neural basis of this kind of knowledge is more or less agreed to depend on a distributed network of cortical brain regions ...


10

Motivation is a massive topic, and it's difficult for me to know what would count as a 'theory' of motivation as it's currently construed. For instance, at one level, we might consider motivation to be the processing of incentive salience on perceived stimuli: you see a cheeseburger, something makes you want it, and so you pursue it. One way of talking ...


10

I'd like to add to Chuck's excellent answer; the computational approach is very well-represented in neuroscience, and actually involves a large number of very heterogeneous methods. Thus, a very different set of neuroscientists and examples have sprung to mind for me. To my mind, the best single example of the utility of a computational approach to ...


10

SVM training is typically done in a batch processing, and thus the order of data presentation doesn't matter. You should consider online learning algorithms, for example, the perceptron learning rule. These algorithms are in general stochastic gradient descent optimization procedures, and easy examples early on with larger learning step would be much more ...


9

There are many neuroscientists who use the techniques of advanced mathematics and statistics to analyze actual neural data for patterns. George Gerstein, who is now retired, has been a pioneer in applying "particle" methods in analyzing neuronal interactions. The originator of the Gravity transform, he used this tool to untangle some of the stochastic ...


9

Very many references may easily be found with a Google search for "mathematical model memory". Probably the most classic and iconic reference is Atkinson and Shiffrin (1965), which is also described on Wikipedia. Its three components and their relationships are nicely encapsulated in this figure: Many other, lesser-known mathematical models of memory ...


8

This should perhaps be a comment, but I don't have the reputation. The other two answers mention that a major drawback to the Bayesian approach is its lack of biological plausibility. However, see for instance: Bayesian inference with probabilistic population codes Ma, W.J. and Beck, J.M. and Latham, P.E. and Pouget, A. Nature Neuroscience, ...


8

I don't know of any NN algorithms that match your definition entirely, and I have looked for them (previously and recently). Here are some papers that I think are close or in the direction that you are exploring. Using theoretical models to analyze neural development (review) An Instruction Language for Self-Construction in the Context of Neural Networks ...


7

I would doubt that the effect you described is an example for the positivity effect, which is a form of attributional bias, rather than connected to basic perceptual phenomena. Instead I think that the experimental results you described can be explained by a priming effect. Priming describes the phenomenon that the exposure to one stimulus influences the ...


7

Dehaene & Changeux (1991) made a neural-network model: The coding units are clusters of neurons organized in layers, or assemblies. A sensonmotor loop enables the network to sort the input cards according to several criteria (color, form, etc.). A higher-level assembly of rule-coding clusters codes for the currently tested rule, which shifts when ...


7

There are several such models in the field of auditory perception. For example Patterson 1996 [1] suggests a model that starts with a simulation of the cochlea and the neural activity and reaches up to perception; Winkler 2006 [2] reviews the process of auditory perception, again from the cochlea up to perception. Somewhat old and does not mention a ...


7

Schroedingers Cat nailed a couple of important points and I need to expand it further. I believe the important roadblock in the creation of an absolutely intelligent machine is the fact that human brain is still not decrypted. There isn't a holistic cognitive model that describes how the brain functions, reacts or take decisions. Even in case such a model ...


6

The use of neural-networks in the cognitive sciences has been around since Turing. However, many of the networks common in connectionism suffer from a lack of biological plausibility. Of these abstract ones, even the ones that try to capture some properties of biological neural networks only do some metaphorically. See for instance the limitations of cascade ...


6

I think that your intuition about the lower "energy ratio" of spurious states explaining their greater susceptibility to unlearning might be correct. In a Hopfield Network spurious states are activity patterns that have not been explicitly embedded in the synaptic matrix, but are nonetheless stable. They are in other words "unwanted" attractor states that, ...


6

Another way of thinking about this is that by progressing easy-to-hard, different intermediate knowledge structures are called into existence in the course of processing. These knowledge structures, built from an agent's encounter with easy problems, can prove useful in its encounter with subsequent and more difficult problems. This idea has been around ...


6

"how to build a system that can read natural language and ... understand it" I am not sure you appreciate what you are wanting here. Just this one piece requires an understanding of the processes of the mind that I don't believe we are at yet. There is a lot of material about how we MIGHT achieve this, but you would need to appreciate the whole cognative ...


6

Another reason for reduced plasticity in adults is that learning something different in the presence of an existing knowledge structure is more difficult than learning from a "blank slate". In a sense, you get interference from the known language (for example). One person who has developed this argument computationally is Jay McClelland in the context of ...


6

Humans actually exhibit both slow and fast learning and they have somewhat different properties. One distinction is between "declarative" memory (for example, facts like "tigers have stripes" or "Paris is the capital of France") and "procedural" learning (such as perceptuo-motor skills like riding a bike or playing a musical instrument). Declarative memory ...


6

As you have already hinted at, the issue is controversial. I could leave it at that and say "no, there is no consensus", and it would be a true answer, but it wouldn't be satisfying, wouldn't it? Instead, I'll briefly define the topic, give a few examples, and then a few recent criticisms. My answer will be weighted somewhat towards "cognition" instead of ...


5

Yes-- you may be very interested in the Human Connectome Project (also here), whose goal is to map human brain connectivity. One of the primary tools used to map functional and anatomical connections is diffusion tensor imaging (DTI). Unlike the more often used T2-weighted images, DTI allows researchers to image white matter tracts directly. The data from ...


5

I am not sure whether the three assumptions on which your question is based are really valid. (1) Why should a high transformation transfer be linked to a high firing rate? Depending on the role of a single neuron within a group, not firing might carry as much as information as firing. (2) Energy conservation might not be linked to the behavior of a ...


5

Here's something I dug up for language: a Computer Science Thesis from Boulder: The Sensorimotor Foundations of Phonology: A Computational Model of Early Childhood Articulatory and Phonetic Development (1994) it discusses what it calls HABLAR (Hierarchical Articulatory Based Language Acquisition by Reinforcement learning). From the reductionist/biology ...


5

Interesting question. I've written up a discussion of the model-based training literature and how it relates structuring task difficulty with practice. That said, I feel it's only a start and my apologies that it is more pitched at cognitive tasks than perceptual tasks. A summary of model-based training systems Fu et al (2006) have a paper on real-time ...


5

Building on @JohnPick's answer and my comment, but being a little bit more formal. The difference can be explained by the difference between parallel versus sequential processing, and the difficulty of the predicate being evaluated. Your specific question is answered by Treisman (1985) (which I summarize in the second section) but I try to provide a more ...


5

To my knowledge, there is no adjusted RMSD. RMSD, unlike $R^2$, isn't typically used to compare models across the literature. $R^2$ represents the proportion of variance explained by the model, a construct which translates well across different experimental designs. Adjusted $R^2$ distorts this by accounting for the number of parameters in your model, but ...


5

I have participated in NIPS, CNS, and COSYNE at least a couple times each. In fact, I have participated in all 3 last year. COSYNE is a smallest conference, but it's growing fast. It's a great conference because it has a good balance between experimentalists and theorists. It takes an extended abstract (2 pages). It emphasizes the systems aspect of the ...



Only top voted, non community-wiki answers of a minimum length are eligible