Take the 2-minute tour ×
Cognitive Sciences Stack Exchange is a question and answer site for practitioners, researchers, and students in cognitive science, psychology, neuroscience, and psychiatry. It's 100% free, no registration required.

Background: In many situations, people use to classify objects without knowing Machine Learning theory. For example, if small children see an unknown animal in the wild, (s)he tries to classify it as dangerous or not dangerous (based on previous "data").

Question: Obviously, humans use any classifier. My question is: Which?

share|improve this question

migrated from stats.stackexchange.com Sep 26 '13 at 19:11

This question came from our site for people interested in statistics, machine learning, data analysis, data mining, and data visualization.

    
We don't have a clue yet. :P –  Memming Sep 27 '13 at 10:26
    
propably none, our "neuronal network" methods are oversimplified and really dumb (from natures perspective), standard machine-"learning" algorithms are also most of the time oversimplifying things too much (baysian bullshit etc.) I think that we need more studies on how the human brain/other more simpler brains (cat/dog/birds,...) do * classification of parts of an object * group/order these parts together to concepts * use the right concept * learn new concepts... There are imho many simple basic principles involved and we must learn more about learning/memory/congition –  Quonux Feb 22 at 18:56
    
@Quonux, I love the "baysian bullshit" phrase. :) –  Anthony Gatlin May 16 at 1:24

2 Answers 2

This is of course a big question and I don't believe that there is a definite answer to it. A very thorough investigation of this matter comes from Rogers and McClelland (2004), who have a developed a parallel distributed theory of acquisition, representation and use of human semantic knowledge. As the name implies, this effort comes from the realm of parallel distributed processing, which is also called connectionism. The books that @DikranMarsupial cited can be seen as the foundations of the field. As far as I understand, the two volumes and their authors (McClelland is only one of them, at least two other important names are David Rumelhart and Geoffrey Hinton) had an enormous impact on the revival of neural networks within cognitive science and on the development of the parallel distributed approach.

Rogers & McClelland (2004) offer an interesting account of how things are grouped into categories, which is what you are asling about. To give a very brief explanation: according to their theory, all objects have certain attributes. For example, a canary has feet, wings and feathers. A robin also has all these things. A pine, on the other hand, has leaves, roots and roots. The same is true for an oak. In reality, there are much more attributes, but this should be enough to make the point.

Objects that tend to covary with one another in a coherent way (because they have similar attributes) are perceived as similar and are classified as belonging to a category, like the canary and the robin. Other objects, that do not covary with those from this first group, but that do covary coherently with other objects, are classified as belonging to another group, like the pine and the oak. It is these coherent categories that are important for classifying objects.

Rogers and McClelland (2004) used a neural network to demonstrate and test their theory. Their findings are consistent with a range of findings from developmental studies. They also have an article, which is basically a short summary of the book (Rogers & McClelland, 2008). If you are interested, you might want to check out this first. The article has peer reviewers commentaries that point out weaknesses of the theory, which is also very interesting.

Since you refered to machine learning in your question: the learning algorithm used in the simulation studies was the backpropagation of error - algorithm. As was also mentioned by @DikranMarsupial, this is not an especially biologically plausible algorithm. Interestingly though, as is also explained in much detail in the book, it is possible to predict the order in which the objects within the training corpus could be classified by an eigenvector decomposition of a special variant of the covariance matrix of the objects. I'm not an expert here, but to me there seems to be a connection to the so called covariance rule (unfortunately I don't have a citation here), which to my knowledge is biologically plausible.

References:

Rogers, T. T, & McClelland, J. L. (2004). Semantic cognition: A parallel distributed processing approach. Cambridge, MA: MIT Press. PDF
Rogers, Timothy T., & McClelland, J. L. (2008). Précis of Semantic Cognition: A Parallel Distributed Processing Approach. Behavioral and Brain Sciences, 31(06), 689. doi:10.1017/S0140525X0800589X PDF

share|improve this answer
    
Yes, very fascinating. Thank you for a fine answer and for including these references. –  Anthony Gatlin May 16 at 1:27

The obvious answer would be neural networks, which is the family of machine learning methods initially inspired by the lowest level structure of the human brain.

See the books by Mclelland et al

Parallel Distributed Processing, Vol. 1: Foundations: Explorations in the Microstructure of Cognition: Foundations v. 1 (Bradford Books)

Parallel Distributed Processing, Vol. 2: Explorations in the Microstructure of Cognition: Psychological and Biological Models v. 2

which give an account of research into neural networks back in the days where biological plausibility was rather more of a consideration that it generally is now (from a machine learning perspective).

share|improve this answer
3  
It is not so obvious; we know that neurons build brain's hardware and work by mixing spike-like signals in a complex way depending on their internal state that likely changes due to the signals they receive. Though the software is still a mystery and certainly none of the ANN ML solution mimics it (as they for instance have separate training and testing phases). In other words, ANNs are like recreating, like, MS Paint by making a very suspicious simulation of a x86 CPU based on its microscope photos and the EM noise it emits. –  mbq Sep 26 '13 at 13:23
    
Of course neural networks are not an exact recreation of the way the human brain works (who uses Hebbian learning these days in neural networks?). However that doesn't mean that there is not useful similarities, such as distributed representations, and the idea that we don't use reason to determine whether an animal is dangerous, but we just "know" it. If any ML algorithm were similar to human cognition, it would be NNs. Of course humans are also capable of reasoning as well, so there is more to it than that. –  Dikran Marsupial Sep 26 '13 at 14:15
1  
The question seems to me to be on the borderline of stats (as it involves ML) as well as cognitive science (which is distinct from ML and ML quite often has no relation to cognitive science, but is essentially computational statistics). Pointing out neural networks is a reasonable direction for Miroslav to investigate (I started with the PDP books, I'll add a link) –  Dikran Marsupial Sep 26 '13 at 14:17
    
@Dikran, "Cognitive Sciences" on Stack Exchange does indeed cover machine learning, the "s" in "Sciences" means we're a multidisciplinary site :-) –  Josh Gitlin Sep 26 '13 at 21:28

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.