Technical Library
|
Support
Introduction to Machine Learning for TRITON AP-DATA
: What happens during training
What happens during training
Machine Learning | TRITON AP-DATA | v8.3.x | 15-Dec-2016
After submitting your examples, the crawler starts going over the files and providing them to the learning algorithms. If the number of files in a folder is very large, a sampling algorithm samples the folder several times and checks for convergence:
If learning is successful (i.e., the data is "learnable"), the following window appears:
By default, the sensitivity level is set to"Default" (an optimal trade-off between false positives (unintended matches) and false negatives (undetected matches)). The training is performed, by default, ignoring outliers, or examples that could be labeled "positive," but that don't seem to belong to the positive set.
You may choose not to ignore the outliers by clicking on "Yes" and changing that to "No."
You may also change the sensitivity level by clicking on the "Default" link, which brings you to this window:
Note that it is important to consider the percentage of unintended and undetected matches before deciding about the sensitivity level. For example, choosing the "Narrow" level in the window shown above will only increase the expected level of undetected matches, without reducing the expected level of unintended matches, and is, therefore, highly undesirable.
Introduction to Machine Learning for TRITON AP-DATA
: What happens during training
Copyright 2016 Forcepoint LLC. All rights reserved.