Skip to content

Commit f02e5a5

Browse files
committed
Updated assignment observations reference
1 parent 2ed0552 commit f02e5a5

File tree

3 files changed

+14
-1
lines changed

3 files changed

+14
-1
lines changed

experiment/assignment.md

Lines changed: 5 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,5 @@
1+
- What are the components of a competitive learning network?
2+
- Explain the difference between pattern clustering and feature mapping.
3+
- What is a self-organizing network?
4+
- What are the salient features of the Kohonen's self-organizing learning algorithm?
5+

experiment/observations.md

Lines changed: 5 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,5 @@
1+
During the course of performance of the experiments, we understand the following:
2+
- The number of iterations help in capturing the pattern in an efficient way. But the network weights generally saturate after some iterations.
3+
- There is no significant change in the weight values of the nodes of CLNN after some iterations.
4+
- There cannot be an accurate capture of pattern, but a better approximation can be reached with more nodes in the network.
5+

experiment/references.md

Lines changed: 4 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1 +1,4 @@
1-
### Link your references in here
1+
- B. Yegnanarayana, Artificial Neural Networks, New Delhi, India : Prentice-Hall of India, p. 201, 1999.
2+
- T. Kohonen, "Self-organized formation of topologically correct feature maps", Biol. Cybernet., vol. 43, pp. 59-69, 1982b.
3+
- T. Kohonen, Self-organization and Associative Memory, 3rd ed., Berlin: Springer-Verlag, 1989.
4+
- T. Kohonen, "Analysis of simple self-organizing process", Biol. Cybernet., vol. 44, pp. 135-140, 1982a.

0 commit comments

Comments
 (0)