Table 7.
Note that N and B are the size of training sets and the number of bins found by the method, respectively. T is the number of iterations required for convergence of the Platt method and M is defined as the total number of models used in the associated ensemble model.
Training Time | Testing Time | |
---|---|---|
Platt | O(NT) | O(1) |
Hist | O(N logN) | O(logB) |
IsoRegC | O(N logN) | O(logB) |
ACP | O(N logN) | O(N) |
ABB | O(N2) | O(N2) |
BBQ | O(N logN) | O(M logN) |
ENIR | O(N logN) | O(M logB) |