Table 3 Resubstitution error and confusion matrix for Random Forest (RF)-100 model
(a) Resubstitution error matrix for 2014 | ||||
|---|---|---|---|---|
Land-use type | Agriculture land | Built-up land | Wasteland | Waterbody |
Agriculture land | 66 | 0 | 2 | 0 |
Built-up land | 0 | 89 | 0 | 1 |
Wasteland | 0 | 0 | 309 | 0 |
Waterbody | 1 | 0 | 1 | 102 |
Training Overall accuracy = 0.99 | ||||
(b) Confusion matrix for 2014 | ||||
|---|---|---|---|---|
Land-use type | Agriculture land | Built-up land | Wasteland | Waterbody |
Agriculture land | 16 | 0 | 14 | 2 |
Built-up land | 1 | 22 | 8 | 9 |
Wasteland | 6 | 2 | 118 | 7 |
Waterbody | 5 | 2 | 10 | 29 |
Testing (Validation) accuracy = 0.74 | ||||
(c) Resubstitution error matrix for 2020 | ||||
|---|---|---|---|---|
Land-use type | Agriculture land | Built-up land | Wasteland | Waterbody |
Agriculture land | 68 | 0 | 1 | 0 |
Built-up land | 0 | 92 | 0 | 1 |
Wasteland | 0 | 0 | 312 | 0 |
Waterbody | 0 | 0 | 3 | 103 |
Training Overall accuracy = 0.99 | ||||
(d) Confusion matrix for 2020 | ||||
|---|---|---|---|---|
Land-use type | Agriculture land | Built-up land | Wasteland | Waterbody |
Agriculture land | 16 | 2 | 11 | 2 |
Built-up land | 1 | 31 | 3 | 2 |
Wasteland | 2 | 3 | 121 | 4 |
Waterbody | 1 | 0 | 19 | 23 |
Testing (Validation) accuracy = 0.79 | ||||