This page describes the input file format for a GenModel. This specification is used by gensvm_read_model() and gensvm_write_model(). The model file is designed to fully reproduce a GenModel.

The model output file follows the format

Output file for GenSVM (version 0.1)
Generated on: Tue Jan 14 12:00:00 2014 (UTC +01:00)

Model:
p = 2.00
lambda = 0.001
kappa = 1.0
epsilon = 1e-06
weight_idx = 1

Data:
filename = /path/to/data_file.txt
n = 150
m = 4
K = 3

Output:
-0.7693429935131153 -1.9335141926875414
+0.3425555992439160 +1.0939198172438194
+0.3100589593140404 +0.9872012663780092
+0.1319873613546321 +0.1207806485439152
+0.8052481376988456 +0.6507524553955120

The first two lines of the file mainly serve a logging purpose, and are ignored when reading the model file. The model section fully describes the model parameters. Next, the data section describes the data file that was used in training and the size of the dataset. Finally, the output section shows the augmented weight matrix GenModel::V, in row-major order.

Todo:: Write kernel specification to model file as well and adjust the format above.