Sciweavers

Free Online Productivity Tools i2Speak i2Symbol i2OCR iTex2Img iWeb2Print iWeb2Shot i2Type iPdf2Split iPdf2Merge i2Bopomofo i2Arabic i2Style i2Image i2PDF iLatex2Rtf Sci2ools

12

ICANN
2009
Springer

favoriteEmaildiscussreport

175views Neural Networks» more ICANN 2009»

Almost Random Projection Machine

13 years 11 months ago

Almost Random Projection Machine

Download www.fizyka.umk.pl

Backpropagation of errors is not only hard to justify from biological perspective but also it fails to solve problems requiring complex logic. A simpler algorithm based on generation and ﬁltering of useful random projections has better biological justiﬁcation, is faster, easier to train and may in practice solve nonseparable problems of higher complexity than typical feedforward neural networks. Estimation of conﬁdence in network decisions is done by visualization of the number of nodes that agree with the ﬁnal decision. Key words: Neural networks, learning, random projections

Wlodzislaw Duch, Tomasz Maszczyk

Real-time Traffic

Artificial Intelligence | ICANN 2009 | Neural Networks | Random Projections | Useful Random Projections |

claim paper

Related Content

» Almost Tight Upper Bound for Finding Fourier Coefficients of Bounded Pseudo Boolean Functi...

» Almost random graphs with simple hash functions

» Experiments with random projections for machine learning

» Random projection algorithms for convex set intersection problems

» Improving Random Projections Using Marginal Information

» Random Forests Feature Selection with KPLS Detecting Ischemia from Magnetocardiograms

» Learning Random Monotone DNF Under the Uniform Distribution

» Random Projection for High Dimensional Data Clustering A Cluster Ensemble Approach

» Random Projection Margins Kernels and FeatureSelection

Post Info
More Details (n/a)

Added	26 May 2010
Updated	26 May 2010
Type	Conference
Year	2009
Where	ICANN
Authors	Wlodzislaw Duch, Tomasz Maszczyk

Comments (0)