AutoML: Binary Prediction Threshold

[Anonymous]

Good afternoon,

Had a quick question with regards to the threshold when applying a binary prediction model through AutoML.

Is the threshold the minimum probability of which the outcome should be defined as "true", or does the threshold represent the cut-off value for the Class 0 (false positive) and Class 1 (false negative) error rate? 

[v-cherch-msft · Accepted solution]

Hi @Anonymous 

Please check if below articles could help.You could read some blogs here.

https://powerbi.microsoft.com/en-us/blog/creating-machine-learning-models-in-power-bi/

https://docs.microsoft.com/en-us/power-bi/service-machine-learning-automated

Regards,

[Anonymous]

Hi @v-cherch-msft 

Thank you for your quick response! Your second link actually answered my question. Thank you so much!