Peter GECZY

Novel Superlinear First Order Algorithms
Peter GECZY  Shiro USUI 
Publication:   IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences
Publication Date: 2004/06/01
Vol. E87-A  No. 6  pp. 1620-1631
Type of Manuscript: PAPER
Category: Neural Networks and Bioengineering
Keyword: 
first order optimization,  steepest descent,  conjugate gradient,  line search subproblem,  classification framework,  neural networks, 
  Summary |  Full Text:PDF (1.1MB)
Superlinear Conjugate Gradient Method with Adaptable Step Length and Constant Momentum Term
Peter GECZY  Shiro USUI 
Publication:   IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences
Publication Date: 2000/11/20
Vol. E83-A  No. 11  pp. 2320-2328
Type of Manuscript: PAPER
Category: Numerical Analysis and Optimization
Keyword: 
first order optimization,  superlinear convergence,  conjugate gradient,  step length,  momentum, 
  Summary |  Full Text:PDF (590.2KB)
Novel First Order Optimization Classification Framework
Peter GECZY  Shiro USUI 
Publication:   IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences
Publication Date: 2000/11/20
Vol. E83-A  No. 11  pp. 2312-2319
Type of Manuscript: PAPER
Category: Numerical Analysis and Optimization
Keyword: 
first order optimization,  convergence rates,  superlinear convergence,  problem descriptor,  algorithm descriptor, 
  Summary |  Full Text:PDF (328.6KB)
Dynamic Sample Selection: Theory
Peter GECZY  Shiro USUI 
Publication:   IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences
Publication Date: 1998/09/20
Vol. E81-A  No. 9  pp. 1931-1939
Type of Manuscript: PAPER
Category: Neural Networks
Keyword: 
dynamic sample selection,  first order optimization techniques,  search direction,  convergence speed, 
  Summary |  Full Text:PDF (701KB)
Dynamic Sample Selection: Implementation
Peter GECZY  Shiro USUI 
Publication:   IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences
Publication Date: 1998/09/20
Vol. E81-A  No. 9  pp. 1940-1947
Type of Manuscript: PAPER
Category: Neural Networks
Keyword: 
dynamic sample selection,  first order optimization techniques,  search direction,  convergence speed, 
  Summary |  Full Text:PDF (735.4KB)