Chris Huyck's Projects
   - Currently the main focus of my research is on recurrent neural networks
       called Cell Assemblies.  There is a lot on this in my open
       CA page.  There are some
       particular problems that I'm interested in and are appropriate
       for MSc students:
   
 - Compare CANT to an
            existing neuronal simulator such as GENESIS.  There is a range of
            possible projects including reimplementing a simple CANT experiment
            using the GENESIS simulator.
 
         - Fixing up the object (OOP) model for CANT.  In particular
             the input class should be able to take advantage of inheritance
             but currently does not.
 
         - Simple Data Mining tasks using CAs, in particular categorisation
         
 
         - Implementing a pixel to line recognition layer
 
   
    I'm also interested in supervising PhD students in this.  Particular
        problems of interest include
         - Variable Binding
 
         - Use of CAs to solve parsing problems
 
         - Input from the environment and interaction with it
 
         - Theoretical Foundations
 
         - Topological Issues
 
         - Datamining using CAs
 
        We usually have funded studentships and I'm willing to discuss
        these projects with anyone who has a reasonable background.
   
    
   - XML as a useful Knowledge Representation Language.  A traditional
       problem of AI has been how to represent knowledge. Frames, semantic
       nets, scripts and other representation formats have been proposed.
       For instance, the semantics or meaning of natural language text 
       is very difficult to specify.  XML provides a new approach to this
       problem; it provides a mechanism for syntactically specificying
       the meaning of a document.  Using XML for a particular domain of
       text can provide an entry to a wide range of queries on a particular
       domain that would function with a very high degree of precision and
       accuracy.  
   
  
   - Text Mining:  I've got some text data that a company is interested
       in getting information from automatically.  I think the best way
       to progress is use my existing parser (Plink) and generate a dictionary
       and output module for the current domain.  Of course this is all 
       flexible.
  
   - A better search engine
   
  
   - Games and AI:  I'm interested in supervising the implementation of
       a bridge game. There may also be scope for a project involving
       interactive video games.
  
   - I have a lego robot on my shelf.  I'd be interested in getting
       it to do interesting things.  If you're interested in writing such
       a thesis, I'd be willing to supervise it.
 
 
 
   
 
 
   
 
This page was developed by 
Chris Huyck 
This page last modified: