Table 1.
Training/validation data sets for each artificial systems A, B and C, with four different ωij intervals.
| ANN training data | ANN validation data | |||||
|---|---|---|---|---|---|---|
| ω−value | A | B | C | A | B | C |
| {-1, 0, +1} | T(A1,ANN) | T(B1,ANN) | T(C1,ANN) | V(A1,ANN) | V(B1,ANN) | V(C1,ANN) |
| {-5, 0, +5} | T(A5,ANN) | T(B5,ANN) | T(C5,ANN) | V(A5,ANN) | V(B5,ANN) | V(C5,ANN) |
| {-10, 0, +10} | T(A10,ANN) | T(B10,ANN) | T(C10,ANN) | V(A10,ANN) | V(B10,ANN) | V(C10,ANN) |
| {-20, 0, +20} | T(A20,ANN) | T(B20,ANN) | T(C20,ANN) | V(A20,ANN) | V(B20,ANN) | V(C20,ANN) |
| Hill training data | Hill validation data | |||||
| A | B | C | A | B | C | |
| {-1, 0, +1} | T(A1,Hill) | T(B1,Hill) | T(C1,Hill) | V(A1,Hill) | V(B1,Hill) | V(C1,Hill) |
| {-5, 0, +5} | T(A5,Hill) | T(B5,Hill) | T(C5,Hill) | V(A5,Hill) | V(B5,Hill) | V(C5,Hill) |
| {-10, 0, +10} | T(A10,Hill) | T(B10,Hill) | T(C10,Hill) | V(A10,Hill) | V(B10,Hill) | V(C10,Hill) |
| {-20, 0, +20} | T(A20,Hill) | T(B20,Hill) | T(C20,Hill) | V(A20,Hill) | V(B20,Hill) | V(C20,Hill) |