I am using Elastic Stack for test log failure analysis. I would like to make a Data Frame Analytics classification job where I create my own classification categories(class labels.) It seems like the tool is currently set up to predict class labels based only on existing fields/terms rather than declaring my own class labels. Is it possible to somehow write my own custom classes and then train the model to classify new entries based on similarity to known class members? Thanks for your time and any guidance you can provide!
I'd consider using the Log Categorization feature of Anomaly Detection: https://www.elastic.co/guide/en/machine-learning/7.9/ml-configuring-categories.html
1 Like
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.