Classification PI model is getting failed in Training Phase

[Vivek Verma]

Dear Team,

I have been working on developing a Noise and Non-Noise PI classification model for our alerts records. However, I have encountered an error during the training process. I would appreciate assistance in understanding and resolving this issue.

NSE0016:Failed to execute component classifierValidator : 
NCE0026:Classification requires at least two classes. 'assignment_group' has only one class

[Brian Bakker · Accepted solution]

@Vivek Verma 

This means the Output field "Assignment Group" on the Input table only has enough records for one Assignment Group, as each class requires a minimum of 30 records to generate a class to make predictions. Even then, if the data is not indicative enough, the Precision will be set to 100, and the Coverage set to 0, so it will never make a prediction on the class.

If you open the Classification solution definition, and open the Filter link, which opens a new tab with the List View of your training dataset for this solution, if you now group by the Output field (i.e. Assignment Group), you can now see how many records there are for each Output value. If < 30 records, it will be ignored, and in this case, it seem you only have one Assignment Group in your training dataset that has > 30 records. You need at least 2 Assignment Groups > 30 records that are also indicative for making a prediction on.

Regards,

Brian

[Vivek Verma]

Thank you so much ❤️

I'm currently working on implementing a feature for the Alert records that will allow for the classification of noise and non-noise alerts. I plan to achieve this by creating a custom field (Choice) on Alert.

My main concern is whether the SN ML model will be capable of identifying a new alert as noise if it matches the patterns from the trained 10k alerts, all of which are classified as noise.

Alternatively, I'm open to exploring other Predictive Intelligence solutions that may be more suitable for this scenario. Additionally, I'm interested in leveraging advanced ML parameters, and based on my research, I believe that XGBoost would be well-suited for precision classification ML.