- Unable to predict. error when running test on Predictive Intelligence solution

FYL
Mega Sage

Hi, I need some help with my setup for PI. Thank you in advance !

When I run a test , it shows me - Unable to predict. message.

FYL_7-1691499245650.png


Here is the background

I have loaded a 13k incident record and created a solution to predict Category based on Short Description. 

FYL_0-1691498860348.png

 

Training was successful but showed the following message
"Low Net Automation (0%), please re-configure and re-train with more data
FYL_1-1691498909679.png

When I look into the Solution statistics, this is what I get. I see 100 Precision for each of the Classes. 

FYL_3-1691499017725.png

FYL_4-1691499028484.png

I would like to update the Precision to 80 at the Solution level assuming that will increase coverage but nothing happens when I click Apply Values 

FYL_5-1691499100531.png

When I try to update it at the class level, it doesn't have other Lookups for me to apply nor can I modify the Estimated precision of a 100.

FYL_6-1691499191441.png

 

 



 

0 REPLIES 0