Sciweavers

280 search results - page 36 / 56
» Multihomogeneous Newton methods
Sort
View
ICONIP
2007
13 years 10 months ago
Principal Component Analysis for Sparse High-Dimensional Data
Abstract. Principal component analysis (PCA) is a widely used technique for data analysis and dimensionality reduction. Eigenvalue decomposition is the standard algorithm for solvi...
Tapani Raiko, Alexander Ilin, Juha Karhunen
IJCAI
1997
13 years 10 months ago
Law Discovery using Neural Networks
This paper proposes a new connectionist approach to numeric law discovery; i.e., neural networks (law-candidates) are trained by using a newly invented second-order learning algor...
Kazumi Saito, Ryohei Nakano
IEICET
2007
131views more  IEICET 2007»
13 years 8 months ago
High Accuracy Fundamental Matrix Computation and Its Performance Evaluation
We compare the convergence performance of different numerical schemes for computing the fundamental matrix from point correspondences over two images. First, we state the problem ...
Ken-ichi Kanatani, Yasuyuki Sugaya
ICPR
2008
IEEE
14 years 3 months ago
Determination of optimal metabolic pathways through a new learning algorithm
In the present article, we introduce a new method for identification of metabolic pathways in constraint based models that consider enzyme and substrate concentrations. It genera...
C. A. Murthy, Mouli Das, Rajat K. De, Subhasis Muk...
IJCNN
2006
IEEE
14 years 2 months ago
On derivation of stagewise second-order backpropagation by invariant imbedding for multi-stage neural-network learning
— We present a simple, intuitive argument based on “invariant imbedding” in the spirit of dynamic programming to derive a stagewise second-order backpropagation (BP) algorith...
Eiji Mizutani, Stuart Dreyfus