The 18th European Conference on Machine Learning (ECML) and the 11th European Conference on Principles and Practice of Knowledge Discovery in Databases (PKDD)

Efficient AUC Optimization for classification

author: Toon Calders, Eindhoven University of Technology

You might be experiencing some problems with Your Video player.

*Slides*
0:00	Efficient AUC Optimization for Classification
2:54	Overview
3:39	Overview - Introduction
3:41	Assessing Classifier Quality (1)
3:59	Assessing Classifier Quality (2)
4:21	Assessing Classifier Quality (3)
4:37	Assessing Classifier Quality (4)
4:59	ROC Curve (1)
5:19	ROC Curve (2)
5:21	ROC Curve (3)
5:23	ROC Curve (4)
5:26	ROC Curve (5)
5:27	ROC Curve (6)
5:28	ROC Curve (7)
5:30	ROC Curve (8)
5:30	ROC Curve (9)
5:37	Area under ROC curve (AUC)
6:31	Properties of AUC
7:10	Overview - Approximating the AUC using polynomials
7:16	Approximating the AUC using polynomials (1)
7:44	Approximating the AUC using polynomials (2)
7:47	Approximating the AUC using polynomials (3)
7:57	Approximating the AUC using polynomials (4)
8:03	Approximating the AUC using polynomials (5)
8:17	Approximating the AUC using polynomials (6)
9:44	Accuracy of approximation - first experiment (1)
10:26	Accuracy of approximation - first experiment (2)
11:26	Approximating the AUC using polynomials
11:59	Accuracy of approximation - first experiment (2)
12:33	Approximating the AUC using polynomials
12:34	Overview - Linear classifier optimizing AUC directly
12:53	Application - maximizing AUC for classification (1)
13:14	Application - maximizing AUC for classification (2)
13:26	Application - maximizing AUC for classification (3)
13:32	Gradient Descent
14:01	Computing the gradient (1)
14:12	Computing the gradient (2)
14:19	Computing the gradient (3)
14:37	Computing the gradient (4)
14:40	Computing values of AUC(f ) along the gradient (1)
14:59	Computing values of AUC(f ) along the gradient (2)
15:10	Computing values of AUC(f ) along the gradient (3)
15:15	Computing values of AUC(f ) along the gradient (4)
15:25	Computing values of AUC(f ) along the gradient (5)
15:48	Computing values of AUC(f ) along the gradient (6)
15:54	Computing values of AUC(f ) along the gradient (7)
15:56	Computing values of AUC(f ) along the gradient (8)
16:52	Computing values of AUC(f ) along the gradient (9)
17:16	Putting it all together (1)
17:21	Putting it all together (2)
17:29	Putting it all together (3)
17:34	Putting it all together (4)
17:42	Overview - Experiments
17:46	Empirical tests - AUC - Approximation vs SVM Perf
18:57	Empirical tests - Time - Approximation vs SVM Perf
19:24	Empirical tests - AUC - Approximation vs Sampling
20:49	Empirical tests - Time - Approximation vs Sampling
21:11	Empirical tests - Skewed class distribution - KDD cup physics dataset
22:51	Overview - Conclusions and Future research
22:54	Conclusions and Future Work
23:19	Future work - Open problems
24:14	Questions?
25:16	- Questions
28:22	- 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.