How to use Splunk Machine Learning toolkit for anomaly detection?

[manish_singh_77]

Hi All,

How to use Splunk Machine Learning toolkit for anomaly detection. There is a use case where we want machine learning to identify the thresholds dynamically based on anomaly detection.

[woodcock]

These should give you what you need:
https://www.splunk.com/en_us/resources/videos/machine-learning-and-anomaly-detection.html
https://docs.splunk.com/Documentation/MLApp/latest/User/DNOlegacyassist
https://answers.splunk.com/answers/594312/detecting-outlier-without-using-splunk-machine-lea.html

[HiroshiSatoh]

”Splunk Machine Learning toolkit” sample data uses data model（CIM) fields.
For now, how about imitating a sample with an actual data model?

[manish_singh_77]

@HiroshiSatoh

I didn't get it, could you please elaborate.

[manish_singh_77]

@HiroshiSatoh

Have you used "Density Function" algorithm for anomaly detection?

[HiroshiSatoh]

The “DensityFunction” algorithm is not used.
Is it an algorithm question? Or is it a question of how to use MLTK?

Is it a case that does not exist in the MLTK sample use case?