Tag Archives: Nesterov’s method

NeSVM (Nesterov’s method for SVM) code for our ICDM 2010 paper

You can now download MATLAB code for NeSVM from here.  In the code, options.mu is a key parameter to adjust the trade-off between consistent decreasing of primal object function, and the speed. So you need to roughly tune it to … Continue reading

Posted in Tianyi's work | Tagged , , | Leave a comment

NESVM: a Fast Gradient Method for Support Vector Machines

This is a work we finished one year ago and is accepted as a regular paper in ICDM 2010. NESVM is a fast SVM algorithm for primal linear and kernel SVM problems, with the optimal convergence rate of first-order method … Continue reading

Posted in Tianyi's work | Tagged , , , , | 2 Comments

Compressed sensing review (1): Reconstruction Algorithms

CS reconstruction algorithms are always the most astonishing thing for people who know compressed sensing at the first time. Because only a few sampling (much less than Shannon-Nyquist sampling rate) can perfectly reconstruct the whole signal is really a big … Continue reading

Posted in notes | Tagged , , , , , , , , | 6 Comments

Fast Gradient Clustering

This is a work we have posted and given a spotlight talk on NIPS 09 Workshop on Discrete Optimization in Machine Learning: Structures, Algorithms and Applications (DISCML). You can find it on: http://www.cs.caltech.edu/~krausea/discml/papers/zhou09fast.pdf Fast Gradient Clustering (FGC) tackles the two … Continue reading

Posted in Tianyi's work | Tagged , , , , , , , | Leave a comment