Newbie question. For the interactive decision tree, how are you deciding which node to split on? Are you favoring the lighter color nodes first or second? Why did you choose the months since last split if it had the second highest logworth? Thanks so much!
Hi: This is an older video from 2015 that features version 13.2 of SAS Enterprise Miner. However, in the video, the narrator explains that he is picking the MONTH_SINCE_LAST_GIFT variable and manually defining a split rule because he says (at time stamp 8:43) that "...We are interested in recent donors. We want to examine the behaviors of people who donated in the last 8 months." The video at this point features the ability to change the split rules based on business knowledge and familiarity with the data. And, since this is the 3rd video in a series of six videos on the topic of Getting Started with SAS Enterprise Miner, the goal in the series is to provide a quick overview on Enterprise Miner without diving into the theory behind each choice. We have a section on Decision Trees in our Applied Analytics using Enterprise Miner class 2.sas.com/6058Kf1ks using the latest version of Enterprise Miner. In addition, with that class as the recommended prerequisite training, we have a separate Decision Tree Modeling class 2.sas.com/6050Kf1kQ which is a 2 day class on Decision Tree Modeling using SAS Enterprise Miner class. So, it's hard to provide a detailed answer on how to set the rules and variable roles in order to run the Decision Tree Node in Enterprise Miner. You may find this documentation page to be useful: 2.sas.com/6052Kf1ka and you may want to post a question regarding Enterprise Miner and Decision Trees in the Data Science forum, here: 2.sas.com/6053Kf1kx . And, here's a great Library article on Decision Tree theory: 2.sas.com/6055Kf1kL that you might find useful.
Hello again, Aoun, and thank you for your patience! Jeff Thompson, one of our Analytical Trainers here at SAS, and a SAS Enterprise Miner expert, has this answer for your question: The 0,1 results are provided by default due to having a binary target. There is nothing the user needs to do to make these available. If in your own results on your computer you are only seeing statistics summarize mean and standard deviation (and not seeing anything about 0,1) it is because you likely have an interval, i.e., numeric, target. Check your metadata and see what measurement level your target is set to.
Thanks for the video. I am trying to build a random forest model using the SAS EMiner and need help on how to implement my final model ie. how can i use the final model to make predictions using fresh data?
Thanks for letting us know that you are using a random forest approach in Enterprise Miner. This video covered the use of random forests in SAS Viya and SAS Visual Analytics. We have another video that illustrates a random forest approach with Enterprise Miner, here: 2.sas.com/6055GuDrX and here's a Community Forum article on the same topic: 2.sas.com/6056GuDrk . We can't track or answer specific usage questions about your data in this space. For assistance using your data in Enterprise Miner, please open a track with SAS Tech Support. To open a track with Tech Support, fill out the form at this link: 2.sas.com/6058GuDrw .
hello, when i clicked the ... button in the interactive it wont show up a new window. is there something wrong with the process before? and how can i fix it?
Sorry to hear you are having trouble, Alya. Were you able to follow the steps in the video up to that point? What occurs when you click the ellipsis? Also, what version of SAS Enterprise Miner are you using?
@@SASSoftware I am also facing the same issue. I got the error "Unable to open Interactive Decision Tree. Please make sure you have run the predecessors to this node and update the path." on the first attempt and then no response at all upon clicking. I have run all the predecessors to this. I have tried restarting the project, re-creating the decision tree node, etc..
Hi there, I'm trying to create a decision tree by myself and just following the video step by step in setting all those parameter in EM. But unfortunately i did not get the first decision tree as you expressed in the video. There was only one knot in my decision tree. I donno why. Would you mind double check it for me? As i was using the sample data from the following website support.sas.com/documentation/onlinedoc/miner/#minerprev. (Example data for SAS Enterprise Miner, 5.2 ZIP) Thanks a lot.!!!
Yang, this video is a set of sequential videos so it would be important to view as a set to make sure you understand all the concepts presented. Let us know if you have had a chance to view the first two? If not, we would recommend that you first start there. In order to have a better understanding of your question, it would help to see an example of your work so we can help better assist you. You may also want to consider joining our SAS Community groups where you can read through other users' posts or post your own question for others to comment on. We highly recommend this forum for all SAS users! See some examples of previous posts on decision trees here 2.sas.com/60558nlOB
Thanks for your quick reply, Actually i have watched the first two episode of the videos, I'm sure i am very familiar with all those concepts. The results turned out in the previous videos were all the same as i did on my local EM. Problem only exists in this part, decision tree. I will try to run it on my laptop this morning to see if there is any default setting in SAS EM. The system should be all the same, right? thanks again for your suggestions and i will look the forum through today. Cheers.
Hi There, The problem doesn't show up on my laptop. Problem seem solved. There might be a mistake i made in importing data. Anyway, I will try it again on my server tonight. FYI Do you konw where can i get the power point of this series video. The contents is so helpful and terrific. Thanks.
Hi there, Just double confirmed on my server, There was an error in my data cleaning process. Sorry for bothering you. Thanks a lot. FYI please tell me where i can get the power point of this series video.
Hi If you could please share the links for previous 2 videos in this series. this would be really handy thank you
+Shivi Bhatia You can view the entire series here th-cam.com/video/489wJm2X0TY/w-d-xo.html
Newbie question. For the interactive decision tree, how are you deciding which node to split on? Are you favoring the lighter color nodes first or second? Why did you choose the months since last split if it had the second highest logworth? Thanks so much!
Thank you for your inquiry! We are looking in to this for you!
Hi:
This is an older video from 2015 that features version 13.2 of SAS Enterprise Miner. However, in the video, the narrator explains that he is picking the MONTH_SINCE_LAST_GIFT variable and manually defining a split rule because he says (at time stamp 8:43) that "...We are interested in recent donors. We want to examine the behaviors of people who donated in the last 8 months." The video at this point features the ability to change the split rules based on business knowledge and familiarity with the data. And, since this is the 3rd video in a series of six videos on the topic of Getting Started with SAS Enterprise Miner, the goal in the series is to provide a quick overview on Enterprise Miner without diving into the theory behind each choice.
We have a section on Decision Trees in our Applied Analytics using Enterprise Miner class 2.sas.com/6058Kf1ks using the latest version of Enterprise Miner. In addition, with that class as the recommended prerequisite training, we have a separate Decision Tree Modeling class 2.sas.com/6050Kf1kQ which is a 2 day class on Decision Tree Modeling using SAS Enterprise Miner class. So, it's hard to provide a detailed answer on how to set the rules and variable roles in order to run the Decision Tree Node in Enterprise Miner. You may find this documentation page to be useful: 2.sas.com/6052Kf1ka and you may want to post a question regarding Enterprise Miner and Decision Trees in the Data Science forum, here: 2.sas.com/6053Kf1kx . And, here's a great Library article on Decision Tree theory: 2.sas.com/6055Kf1kL that you might find useful.
Hey there, quick question how did you bring the “0”,“1”% into the node statistics? I only see an option for Average and Standard Deviation
Aoun, thank you for your inquiry! We are checking on this for you!
Hello again, Aoun, and thank you for your patience! Jeff Thompson, one of our Analytical Trainers here at SAS, and a SAS Enterprise Miner expert, has this answer for your question:
The 0,1 results are provided by default due to having a binary target. There is nothing the user needs to do to make these available. If in your own results on your computer you are only seeing statistics summarize mean and standard deviation (and not seeing anything about 0,1) it is because you likely have an interval, i.e., numeric, target. Check your metadata and see what measurement level your target is set to.
Thanks for the video. I am trying to build a random forest model using the SAS EMiner and need help on how to implement my final model ie. how can i use the final model to make predictions using fresh data?
Thanks for letting us know that you are using a random forest approach in Enterprise Miner. This video covered the use of random forests in SAS Viya and SAS Visual Analytics. We have another video that illustrates a random forest approach with Enterprise Miner, here: 2.sas.com/6055GuDrX and here's a Community Forum article on the same topic: 2.sas.com/6056GuDrk .
We can't track or answer specific usage questions about your data in this space. For assistance using your data in Enterprise Miner, please open a track with SAS Tech Support. To open a track with Tech Support, fill out the form at this link: 2.sas.com/6058GuDrw .
hello, when i clicked the ... button in the interactive it wont show up a new window. is there something wrong with the process before? and how can i fix it?
Sorry to hear you are having trouble, Alya. Were you able to follow the steps in the video up to that point? What occurs when you click the ellipsis? Also, what version of SAS Enterprise Miner are you using?
@@SASSoftware Same problem. SAS Enterprise Miner 15.1
I’m having the exact same issue.
Everything worked to that point.
same here too everything work till this step
@@SASSoftware I am also facing the same issue. I got the error "Unable to open Interactive Decision Tree.
Please make sure you have run the predecessors to this node and update the path." on the first attempt and then no response at all upon clicking.
I have run all the predecessors to this. I have tried restarting the project, re-creating the decision tree node, etc..
Hi there, I'm trying to create a decision tree by myself and just following the video step by step in setting all those parameter in EM. But unfortunately i did not get the first decision tree as you expressed in the video. There was only one knot in my decision tree. I donno why. Would you mind double check it for me? As i was using the sample data from the following website support.sas.com/documentation/onlinedoc/miner/#minerprev. (Example data for SAS Enterprise Miner, 5.2 ZIP) Thanks a lot.!!!
Yang, thank you for your inquiry! We are currently looking into this for you.
Yang, this video is a set of sequential videos so it would be important to view as a set to make sure you understand all the concepts presented. Let us know if you have had a chance to view the first two? If not, we would recommend that you first start there. In order to have a better understanding of your question, it would help to see an example of your work so we can help better assist you. You may also want to consider joining our SAS Community groups where you can read through other users' posts or post your own question for others to comment on. We highly recommend this forum for all SAS users! See some examples of previous posts on decision trees here 2.sas.com/60558nlOB
Thanks for your quick reply, Actually i have watched the first two episode of the videos, I'm sure i am very familiar with all those concepts. The results turned out in the previous videos were all the same as i did on my local EM. Problem only exists in this part, decision tree. I will try to run it on my laptop this morning to see if there is any default setting in SAS EM. The system should be all the same, right? thanks again for your suggestions and i will look the forum through today. Cheers.
Hi There, The problem doesn't show up on my laptop. Problem seem solved. There might be a mistake i made in importing data. Anyway, I will try it again on my server tonight. FYI Do you konw where can i get the power point of this series video. The contents is so helpful and terrific. Thanks.
Hi there, Just double confirmed on my server, There was an error in my data cleaning process. Sorry for bothering you. Thanks a lot. FYI please tell me where i can get the power point of this series video.
Having to use a GUI to build a model - Yuck!