(One intermediate revision by the same user not shown)
Line 11: Line 11:
 
Today we finished discussing the the Parzen window method for estimating the probability density function at a point x of the feature space using samples. In particular, we discussed how, in the context of a decision problem, this technique boils down to a majority vote among the neighboring samples. However, it was pointed out that using different window volumes for different classes might improve the result of this voting procedure. Also, en interesting connection to the sampling theorem was pointed out by a student.  
 
Today we finished discussing the the Parzen window method for estimating the probability density function at a point x of the feature space using samples. In particular, we discussed how, in the context of a decision problem, this technique boils down to a majority vote among the neighboring samples. However, it was pointed out that using different window volumes for different classes might improve the result of this voting procedure. Also, en interesting connection to the sampling theorem was pointed out by a student.  
  
Note that the [[Hw2_ECE662_S12|second homework assignment]] is not posted. See you after Spring break!
+
Note that the [[Hw2_ECE662_S12|second homework assignment]] is now posted. See you after Spring break!
  
 
==Related Rhea Pages==
 
==Related Rhea Pages==
 
*[[Hw2_ECE662_S12|HW2 ECE662 Spring 2012]]
 
*[[Hw2_ECE662_S12|HW2 ECE662 Spring 2012]]
 
*[[ECE662_hw2_discussions|HW2 discussion from ECE662 Spring 2010]]
 
*[[ECE662_hw2_discussions|HW2 discussion from ECE662 Spring 2010]]
*[[Lecture_16_-_Parzen_Window_Method_and_K-nearest_Neighbor_Density_Estimate_Old_Kiwi|Lecture 16, ECE662, Spring 2008]]
+
*[[Lecture_16_-_Parzen_Window_Method_and_K-nearest_Neighbor_Density_Estimate_OldKiwi|Lecture 16, ECE662, Spring 2008]]
  
 
Previous: [[Lecture17ECE662S12|Lecture 17]]
 
Previous: [[Lecture17ECE662S12|Lecture 17]]

Latest revision as of 10:34, 22 March 2012


Lecture 18 Blog, ECE662 Spring 2012, Prof. Boutin

Thursday March 8, 2012 (Week 9)


Quick link to lecture blogs: 1|2|3|4|5|6|7|8| 9|10|11|12|13|14|15|16|17|18|19|20|21|22|23|24|25|26|27|28|29|30


Today we finished discussing the the Parzen window method for estimating the probability density function at a point x of the feature space using samples. In particular, we discussed how, in the context of a decision problem, this technique boils down to a majority vote among the neighboring samples. However, it was pointed out that using different window volumes for different classes might improve the result of this voting procedure. Also, en interesting connection to the sampling theorem was pointed out by a student.

Note that the second homework assignment is now posted. See you after Spring break!

Related Rhea Pages

Previous: Lecture 17

Next: Lecture 19


Comments

Please write your comments and questions below.

  • Write a comment here
  • Write another comment here.

Back to ECE662 Spring 2012

Alumni Liaison

Correspondence Chess Grandmaster and Purdue Alumni

Prof. Dan Fleetwood