A Constrained Optimization Approach to Preserving Prior Knowledge During Incremental Training

13 years 10 months ago

Download fred.mems.duke.edu

In this paper, a supervised neural network training technique based on constrained optimization is developed for preserving prior knowledge of an input

Silvia Ferrari, Mark Jensenius

Real-time Traffic

Equality Constraints | Incremental Training | Prior Knowledge | TNN 2008 |

claim paper

» Adaptive Feedback Control by Constrained Approximate Dynamic Programming

» 3D KnowledgeBased Segmentation Using PoseInvariant HigherOrder Graphs

» Decision Theoretic Modeling of Human Facial Displays

» Monocular Tracking 3D People By Gaussian Process SpatioTemporal Variable Model

» Support vector regression from simulation data and few experimental samples

» Example Based Nonrigid Shape Detection

» Optimal minimumdelayarea zeroskew clock tree wiresizing in pseudopolynomial time

Post Info
More Details (n/a)

Added	15 Dec 2010
Updated	15 Dec 2010
Type	Journal
Year	2008
Where	TNN
Authors	Silvia Ferrari, Mark Jensenius

Comments (0)

Sciweavers

A Constrained Optimization Approach to Preserving Prior Knowledge During Incremental Training

Equality Constraints | Incremental Training | Prior Knowledge | TNN 2008 |

Explore & Download

Productivity Tools

Document Tools

Image Tools

Sciweavers