en-es
en
0.25
0.5
0.75
1.25
1.5
1.75
2
Divide and Conquer Kernel Ridge Regression
Published on Sep 02, 20134158 Views
We study a decomposition-based scalable approach to performing kernel ridge regression. The method is simply described: it randomly partitions a dataset of size N into m subsets of equal size, compute
Related categories
Chapter list
Divide and Conquer Kernel Ridge Regression00:00
Problem set-up00:46
Kernel regression review - 101:31
Kernel regression review - 202:16
Kernel ridge regression 03:24
Think about large datasets - 104:21
Think about large datasets - 205:26
Our main idea06:30
Fast Kernel Ridge Regression (Fast-KRR)07:55
Theoretical result10:10
Apply to specific kernels - 112:01
Apply to specific kernels - 213:33
Simulation study14:20
Compare Fast-KRR and exact KRR14:43
Threshold for data partitioning15:34
Summary16:25
Open problems17:07