A consultant plans to extract an Opportunity object for training and wants to create a story to identify opportunities that are likely to be won. Which two actions should the consultant take to prepare the training dataset? (Choose two.)
A.
Exclude IsClosed (FALSE).
B.
Maximize on IsWon (TRUE).
C.
Include opportunities that are in all stages for completeness.
D.
Include StageName because this field is highly correlated to IsWon (True).
@Mangoex, I disagree.
Per the question:
"A consultant plans to extract an Opportunity object for training and wants to create a story to identify opportunities that are likely to be won."
identify opportunities that are LIKELY to be won
As such we need to draw a comparison between those that have been won with those that are still in a cycle (not lost).
B and C are the correct answer
A: you need records with outcome, unclosed records do not have any outcome; B: you need to maximize isWon for model to know, which value are you interested in; C: is the opposite of A, contains lot of records with no meaning for model D: highly correlated values are counterproductive
Agree with bshreeves. B and C.
MangoEx logic is flawed. You can’t train a model to maximise an expected outcome by ONLY supplying it with data where the outcome is already maximised. You need a comprehensive dataset that cuts across multiple scenarios so that the model can learn what drives the intended outcome.
A and B are the correct answers.
We are only interested in opportunities that have already reached the outcome. For Opportunities, it typically means when the Opportunity is closed (IsClosed= true), so we do not need isClosed = false. You may also consider filtering for a specific segment of your data as insights that drive faster sales for your enterprise customers may be completely different from your small business customers.
We are not using StageName, because this field designates the sales process stage. Since we are training when IsWon=true, the model will only know the stage “Closed Won”. When it sees stages like “Qualification”, for example, the model won’t know what to do with that value and consider it out of bounds for the model.
https://www.salesforceblogger.com/2020/03/11/preparing-your-data-for-einstein-discovery/
A voting comment increases the vote count for the chosen answer by one.
Upvoting a comment with a selected answer will also increase the vote count towards that answer by one.
So if you see a comment that you already agree with, you can upvote it instead of posting a new comment.
bshreeves
Highly Voted 1Â year, 5Â months ago840acea
Most Recent 1Â week, 4Â days agohananetest
3Â months, 1Â week agoRMEZZA
6Â months, 1Â week agocarbonhero
10Â months, 1Â week agopesiqueira
11Â months, 4Â weeks agofacemelter2000
1Â year, 5Â months agoMangoex
1Â year, 9Â months agosnusmumrick
1Â year, 10Â months ago