Output of pnd application. |
Output format of the application for training and testing Parallel Non linear Dichotomizers
The application print on standard output the following informations:
See also pnd
- PND type
- General features of dichotomizers: number of dichotomizers, hidden layers, hidden units, inputs
- Type of learning algorithm, its parameters and stop conditions.
- Training results of dichotomizers: RMS error and iterations
- Testing results: Number of errors and error rate, confusion matrix
- Elapsed CPU time
Example:
This example shows the output of and OPC-PND for a 6 classes classification problem:OPC-PND Number of classes = 6 Number of dichotomizers = 6 Minimum Hamming distance between codewords = 2 Dichotomizers specifications: Number of layers: 2 Input dimension: 3 Units of the hidden layer: 7 Stop conditions: Maximum number of iterations : 300 RMS error threshold : 0.1 Learning algorithm: Alg. Gradient descent Learning rate : 0.03 By pattern learning. Dichotomizer n. 1 Iterations : 11 RMS Error: 0.09812 Dichotomizer n. 2 Iterations : 13 RMS Error: 0.09847 Dichotomizer n. 3 Iterations : 14 RMS Error: 0.09754 Dichotomizer n. 4 Iterations : 11 RMS Error: 0.0866 Dichotomizer n. 5 Iterations : 15 RMS Error: 0.09934 Dichotomizer n. 6 Iterations : 13 RMS Error: 0.09988 Confusion matrix : 199 0 0 0 0 0 1 200 0 0 0 0 0 0 200 1 0 0 0 0 0 199 0 0 0 0 0 0 200 0 0 0 0 0 0 200 PND errors: Number of errors : 2 : Error rate % : 0.1667 : CPU time (min.sec) : 0.24 :
Alphabetic index Hierarchy of classes