Alireza Nejati

Questions and Answers (41) View all

• Answer added in Discrete Event Systems
  3 Do you agree with that Artificial Neural Networks are able to approximate discrete functions?
  By MengChang Wang · Huazhong University of Science and Technology
  Alireza Nejati · University of Auckland

  The Cybenko theorem clearly states that ANNs have the ability to learn any function, including discrete ones. Whether ANNs are the most efficient way... [more]

  The Cybenko theorem clearly states that ANNs have the ability to learn any function, including discrete ones. Whether ANNs are the most efficient way of doing this is a different matter entirely.
  Following
• Answer added in Pattern Recognition
  1 What's the difference between sparse representation, sparse coding and sparse auto encoder?
  By Fuqiang Chen · Tongji University
  Alireza Nejati · University of Auckland

  Sparse coding: representation of a signal family in terms of a code that usually ends up zero or close to zero in most of its elements. Sparse repres... [more]

  Sparse coding: representation of a signal family in terms of a code that usually ends up zero or close to zero in most of its elements. Sparse representation: the representation of a single element of this family in terms of the sparse code. Sparse auto-encoder: a unsupervised machine learning algorithm that 'learns' a sparse code for a signal family.
  Following
• Answer added in Machine Learning
  18 Problem with backpropagation.
  By Prashant Chavan
  Alireza Nejati · University of Auckland

  I think you need to provide some more information i.e. the ANN software you're using (is it the matlab toolbox?) and how you're training your net. Als... [more]

  I think you need to provide some more information i.e. the ANN software you're using (is it the matlab toolbox?) and how you're training your net. Also, what training algorithm you're using (LM, GDM, etc.)
  Following
• Answer added in Control Theory
  6 There is a relationship between knot theory and control theory or dynamic systems?
  By Jessica Mendiola · Instituto Potosino de Investigación Científica y Tecnológica
  Alireza Nejati · University of Auckland

  Are you referring to the fact that periodic orbits of certain dynamical systems can be considered as being knots? 

  Are you referring to the fact that periodic orbits of certain dynamical systems can be considered as being knots?
  Following
• Answer added in Pattern Recognition
  4 How can mathematically prove that increasing entropy lead to decreasing over fitting?
  By Arian Mosallanezhad · Khaje Nasir Toosi University of Technology
  Alireza Nejati · University of Auckland

  Look at the Bayesian information criterion (BIC): http://en.wikipedia.org/wiki/Bayesian_information_criterion 

  Look at the Bayesian information criterion (BIC): http://en.wikipedia.org/wiki/Bayesian_information_criterion
  Following