Since this was the only method that could be used for establishing a cut-off value for bison and caribou sera, sensitivity and specificity estimates could not be generated for this ELISA test for these species.
For each of the folds, probabilities are generated for the test compounds after training.
The test sets generated for this model have the following parameters: (RP_1): For each of the eight inputs of the Duplex model two random values were selected in the range of ([-1000 [-1000) (valid inputs) and no values were selected outside of the domain input because no further improvement could be achieved.
The test sets generated for this model have the following parameters: (RP_1): For each of the three inputs of the Quadratic model two random values were selected in the range of ([-100, 100]) (valid inputs), and two were selected in the range of ([-1000 [-1000cup ]100, 1000]) (invalid inputs).
The test sets generated for this model have the following parameters: (RP_1): For each of the three inputs of the Tiny model two random values were selected in the range of ([-100, 100]) (valid inputs), and two were selected in the range of ([-1000 [-1000cup ]100, 1000]) (invalid inputs).
The test sets generated for this model have the following parameters: (RP_1): For the temperature input of the Flow Control model two random values were selected in the range of ([-100, 300]) (valid inputs), and two were selected in the range of ([-1000 [-1000cup ]300, 1000]) (invalid inputs).
The domain of the first three inputs is from 10 to 30 and the threshold input domain is 2. The test sets generated for this model have the following parameters: (RP_1): For each of the first three inputs of the Generic Voter model two random values were selected in the range of ([-10, 30]) (valid inputs), and two were selected in the range of ([-1000 [-1000up ]30, 1000]) (invalid inputs).
Two test statistics were generated for each model tested: a t-test on the multiplicative interaction term (Pixn), and an F-test on the overall model (Pmod).
