Research Article
A Comparative Study of Land Cover Classification by Using Multispectral and Texture Data
Table 5
NDA architecture for multispectral dataset.
| Input layers = 5 | 1st hidden layer = 5 | 2nd hidden layer = 2 |
| Learning rate eta = 0.20 | Back propagation iteration = 200000 | Optimized iteration limit = 70 |
| Output layers = 5 |
|
|