Paging Saitama and Xavier.... LOLs. Anyway, I think "Bald" would be a human guy with no hair on top of his head, and "No Hair" would be like robots, aliens, mutant crocodiles, etc.
I can't wait to start using this for educational environment. I visualize this helping me identify which apartment complex has the most students who need interventions for attendance. Lot's of ways to get there but this seems really quick.
First thing coming to my mind is to use this during our desition making process, where we have a desition matrix when qualify and quantify desition based on facts (range between 1 to 10) and here I thing Power Bi can help us to make visual the final desition. Nice!!!
Definitely show it off, but I think it's important to make a distinction that this is AI, not machine learning. Machine learning implies that the model gets better as time goes by and more information is fed into it. I THINK what's happening here is some mix of different statistical models happening at the same time and the visual picks the best fitting.
Nice, Nice, thank you for the movie of Power BI Key Influencer visual, real-scenario and common question. 😆😆😆 I hope movie more about this visual, to learn this visual more, share this visual more, achieve more
This will work with calculated columns and calculated tables? Because if works I could insert my measurements in my calculated table just using DAX without using M.
Yes as of next month we will work with both numerical and categorical target columns (including calculated columns and tables). We are hoping users will use this as a workaround until measure support comes :)
Hey Patrick, could you explain more about what is going on in the background of this visual? Is there one or several models that are being tested here? What makes this intelligent vs the 'dumb' visualizations? My clients are asking what kind of modeling is happening in the back end before they really start trusting it.
Hi Luis please check out the documentation here and scroll down all the way to the end. We have a section about how we calculate influencers and segments. As a summary - we run a logistic regression (p-value of 0.05) for the influencers and decision tree (also p-value of 0.05) for segments. This gives us the ranked list of influencers and segments docs.microsoft.com/en-us/power-bi/visuals/power-bi-visualization-influencers
1:24 - What does Patric mean when he asks about "a measure" when she says "a metric", aren't measures the metrics we normally report on? what's the difference between a measure and a metric?
Hey Patrick! Great content, as always! Just discovered this visual and it's BANANAS!! :D A new world of possibilities would be opened to me if I could just retrieve the Key influencers results as text, do you know if there's something like a REST connector or are they stored anywhere accesible?? Thanks :)
Thanks. That is cool and looks useful for some of the data I deal with. On a simpler note, is there or are you working on a visualization for a traditional Pareto chart? I know I can build one using the existing visualizations and some additional calcs but it would be nice if there was one already available.
Didn't get what this all about. What does it mean 1.36x ? Is it some sort of probability that bald guy is 1.36 times more likely (than who ?) to be also a bad guy ?
I have requirement to put drop down filter inside the pie chart. For example I have three type of dimensions City State Country I have a chart which is showing data by default on city basis on pie chart. But My customer wants to add one drop down on top of pie chart and show three option inside the drop down as mentioned above. So when I change value of drop down from city to state same pie chart should be shown by state and if I change drop down value to Country then my pie chart show distribution of my data per country. Is that possible in power BI? Thanks in advance
Open any Power BI file. File/Options&Settings/Options/Preview features. Check "Key Influencers" and you are good to go. Of course download the last update of the PBI if you don't have it already. Hope this works.
What is the AI behind? It looks like quite basic statistics algorithms packed in nice UI. Do not take me wrong. Big thumbs up for it, but this is not AI.....
Hi @@Enties you are correct, we are running a logistic regression and decision tree. If you are defining AI in the sense of having a feedback loop where the system becomes more intelligent with time- yes you are absolutely correct we are not doing that with the visual. The visual is using some well known statistical models to help business users find insights more quickly in their data :)