|
|||||
|
|
On the combination of supervised and unsupervised learning |
|
| Authors: | Nathan Intrator |
| Institution: | School of Mathematical Sciences, Tel-Aviv University, Ramat-Aviv 69978, Israel |
| Abstract: | The bias/variance dilemma is addressed in the context of neural networks. A bias constraint based on prior knowledge about the underlying distribution of the data is discussed as a means for reducing the overall error measure of a classifier. |
| Keywords: | |
| 本文献已被 ScienceDirect 等数据库收录！ | |