Breaking news from around the world
Get the Bing + MSN extension
Now Available in Community - MBAS 2019 Presentation Videos
Catch the most popular sessions on demand and learn how Dynamics 365, Power BI, PowerApps, Microsoft Flow, and Excel are powering major transformations around the globe. | View Gallery
2019 release wave 2 Discover the latest updates to Dynamics 365Release overview guides and videos Release Plan | Early Access Availability
Ace your Dynamics 365 deployment with packaged services delivered by expert consultants. | Explore service offerings
Connect with the ISV success team on the latest roadmap, developer tool for AppSource certification, and ISV community engagements | ISV self-service portal
The FastTrack program is designed to help you accelerate your Dynamics 365 deployment with confidence.
FastTrack Program | Finance TechTalks | Customer Engagement TechTalks | Talent TechTalks | Upcoming TechTalks
Dynamics 365 AI for Sales creates models for scoring the probability of qualifying a lead/ an opportunity. The scores are then bucketed into 4 Grades A to D. In the configuration screen it is possible to define the range of scores for each grade bucket.
My question now pertains the accuracy metric above that changes when you change the bucket boundaries. If all test samples are put in the D-Bucket then its conversion rate is equal the total conversion rate and equal the accuracy. If all samples are put in a "higher" bucket then the conversion rate of that bucket is shown higher than the total conversion rate. Also the accuracy rises. Being placed in a certain bucket apparently has an influence over the accuracy score.
Can somebody explain how the accuracy is computed. And also what the influence of the grading is.
You could refer to this post:
no sorry. This post seems to be only concerned with how the actual scores for leads and opportunities are computed. I am interested in interpreting the summary page in the configuration and its metrics. See here or here for a view of the summary screen.
PS: oh yeah, I am especially interested in the green "prediction accuracy" values which as I said earlier changes with the bucket boundaries.
For the 1st question, yes you can define your own range of scores for each bucket by updating the score ranges and then saving the model using the Save button.
Accuracy is different from the Grades that are generated from the prediction scores. Accuracy of a model is calculated on a training data set which is a subset of data available on the CRM Org. Grading however classifies the scores that are generated every day (depending on the model that was generated at training time) for all open leads/opportunies into the different buckets.
Business Applications communities