NIPS Workshop on New Challenges in Theoretical Machine Learning: Learning with Data-dependent Concept Spaces, Whistler 2008

Semi-Supervised Learning and Learning via Similarity Functions: two key settings for Data-dependent Concept Spaces

author: Avrim Blum, School of Computer Science, Carnegie Mellon University
published: Dec. 20, 2008, recorded: December 2008, views: 536

Slides

Slides
0:00	Semi-Supervised Learning and Learning via Similarity Functions: Two key settings for Data-Dependent Concept Spaces
0:05	Theme of the workshop: Data-Dependent Concept Spaces - 1
1:02	Theme of the workshop: Data-Dependent Concept Spaces - 2
1:18	Semi-Supervised Learning - 1
1:47	Semi-Supervised Learning - 2
2:24	Ex 1: Co-Training - 1
3:12	Ex 1: Co-Training - 2
3:34	Ex 1: Co-Training - 3
4:11	Simple Example: Intervals
5:14	Ex 1: Co-Training - 4
5:35	Ex 1: Co-Training - 5
5:48	Ex 2: S3VM [Joachims98] - 1
6:44	Ex 2: S3VM [Joachims98] - 2
7:05	Ex 3: Graph-Based Methods - 1
8:04	Ex 3: Graph-Based Methods - 2
8:15	Ex 3: Graph-Based Methods - 3
8:44	Ex 3: Graph-Based Methods - 4
8:52	Several Different Approaches
9:10	Semi-Supervised Model - 1
9:35	Semi-Supervised Model - 2
11:24	Semi-Supervised Model - 3
12:17	Formally - 1
13:54	Formally - 2
14:17	Can Use to Prove Sample Bounds - 1
16:35	Can Use to Prove Sample Bounds - 2
16:52	Can Use to Prove Sample Bounds - 3
17:24	Can Use to Prove Sample Bounds - 4
20:23	Examples
20:39	- Questions
27:22	Conclusions part 1
27:50	Topic 2: Learning with Similarity Functions
27:54	Quick Reminder on Kernels
28:45	Moreover, Generalize Well if Good Margin - 1
29:04	Moreover, Generalize Well if Good Margin - 2
29:42	Can We Define a Notion of a Good Measure of Similarity that…
30:28	Warmup Property
31:30	Just Do Average-Nearest-Nbr
31:54	But not Broad Enough - 1
32:43	But not Broad Enough - 2
33:08	Broader Defn… - 1
34:04	Broader Defn… - 2
34:17	How to Use such a Sim Fn? - 1
36:32	How to Use such a Sim Fn? - 2

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.

Lecture popularity: You need to login to cast your vote.