Training Support Vector Machines: Status and Challenges

author: Chih-Jen Lin, National Taiwan University

You might be experiencing some problems with Your Video player.

*Slides*
0:00	Training Support Vector Machines: Status and Challenges
0:14	Outline
1:08	Outline - Introduction to SVM
1:11	Support Vector Classification
2:37	Support Vector Classification (Cont’d)
2:41	Support Vector Classification
2:56	Support Vector Classification (Cont’d)
4:50	Outline - Solving SVM Quadratic Programming Problem
4:51	Large Dense Quadratic Programming
6:09	Decomposition Methods
7:27	Avoid Memory Problems
9:03	How Decomposition Methods Perform? - 1
10:51	How Decomposition Methods Perform? - 2
11:51	Issues of Decomposition Methods
13:07	Outline - Training Large-Scale Data
13:08	Parallel: Multi-Core/Shared Memory
16:18	Parallel: Distributed Environments
18:25	Approximations
20:03	Approximations (Cont’d)
20:49	Parallelization or Approximation
21:37	Parallelization or Approximation (Cont’d)
22:38	Approximations (Cont’d)
22:45	Parallelization or Approximation (Cont’d)
23:18	Outline - Linear SVM
23:23	Linear Support Vector Machines
24:35	Linear Support Vector Machines (Cont’d)
24:45	Linear SVM for Large Document Sets
25:07	Linear SVM
25:41	A Comparison: LIBSVM and LIBLINEAR
26:32	Revisit Decomposition Methods - 1
28:01	Revisit Decomposition Methods - 2
28:43	Testing Accuracy (Time in Seconds)
28:44	Analysis
30:24	Analysis (Cont’d) - 1
31:14	Analysis
31:55	Analysis (Cont’d) - 1
31:57	Analysis (Cont’d) - 2
32:54	Outline - Discussion and Conclusions
32:56	Discussion and Conclusions
34:40	Discussion and Conclusions (Cont’d)
36:17	- Questions
36:39	- Questions
37:28	- Questions
38:16	- Questions

Lecture rating

People found this lecture:
Worth seeing
because it is:
Valuable and informative
Well presented
Easily understandable
Acceptably recorded
You need to login to cast your vote.

Report a problem or upload files

If you have found a problem with this lecture or would like to send us extra material, articles, exercises, etc., please use our ticket system to describe your request and upload the data.
Enter your e-mail into the 'Cc' field, and we will keep you updated with your request's status.