About WendyCzika

WendyCzika · ‎03-20-2016

Can you do a Find of "ERROR:" in the the log from the Regression node and paste in that section?

WendyCzika · ‎03-17-2016

You can fit a GLM with the zero-inflated Poisson distribution in the HP GLM node in Enterprise Miner (in releases 13.1 and beyond). But that's for a target that has many 0's. For inputs that are skewed, you could still use a Log transformation, just need to add a constant to the variables first to be able to log 0. The Transform Variables node in EM can do the log transformation and will automatically add a constant.

WendyCzika · ‎03-15-2016

You can try tweaking some of the options for growing the tree to be less restrictive: for example, lowering the values for the properties Minimum Categorical Size or Leaf Size, or raising the value for Significance Level if using the ProbF or ProbChisq splitting criteria.

WendyCzika · ‎03-08-2016

I will look into it, just need you to clarify whether it is the KS in the Fit Statistics table or in the KS table used in the KS Plot? Thanks! Wendy Czika SAS Enterprise Miner R&D

WendyCzika · ‎03-03-2016

You need to use the Metadata node following the Neural Network node to change the roles of the observed target to REJECTED and the predicted target (posterior probability for the event level if a nominal target) to TARGET. Then use a Decision Tree node after that.

WendyCzika · ‎03-02-2016

Neural networks don't directly give you variable importance, and all of the inputs are included in the model (no selection is done, so there is no reduced set). But see this post about a technique of using a decision tree as a surrogate model for calculating variable importance based on the neural network model that was fit: https://communities.sas.com/t5/SAS-Data-Mining/Interpreting-Neural-Network/m-p/250372/highlight/true#M3705

WendyCzika · ‎02-29-2016

I'm guessing it's the HP Forest node that is causing the error. Since that node does not produce DATA step score code, it cannot be used with the Ensemble node.

WendyCzika · ‎02-23-2016

The t-value is the parameter estimate (aka coefficient) divided by its standard error. The significance of this statistic based on the T distribution is given by the P Value column, so the effects with the smallest p-values are the most significant.

WendyCzika · ‎02-18-2016

Here are a couple resources on survival data mining. http://support.sas.com/rnd/app/video/index.html#txtmine (See videos Introduction to Survival Data Mining and New Features in the SAS Enterprise Miner 12.3 Survival Node) and this tip that gives a link to an example SAS Enterprise Miner template and accompanying documentation for survival data mining: https://communities.sas.com/t5/SAS-Communities-Library/Survival-Data-Mining-by-Example-in-SAS-Enterprise-Miner/ta-p/231023 https://support.sas.com/resources/papers/proceedings12/132-2012.pdf

WendyCzika · ‎02-16-2016

A common approach is to train a decision tree using the predicted values from a neural network to get measures of variable importance for your inputs. This decision tree is then essentially a surrogate model that acts as a proxy to the complex logic of the neural network. To do this in EM, you would attach a Metadata node after your Neural Network node and set the role to 'Target' for either: - one of the columns of posterior probabilities from your Neural Network node, e.g. P_TargetEvent where Target is the name of your original binary target and Event is the event level of the target, - the predicted target, I_Target The original target should be given the role 'Rejected' now. Then run a Decision Tree to see what variables are important in the model. You can similarly get variable importance by including a Model Comparison or Score node after the Neural Network node in your flow, then a Reporter node at the end of the flow with the Nodes property set to Summary. This also uses a decision tree to calculate variable importance. Hope that helps, Wendy

WendyCzika · ‎02-16-2016

I believe the correct names of the data sets are actually TreeZ_outseq, TreeZ_outnodes, and TreeZ_outstats (assuming that is what corresponds to statsbynode). You can see all the data sets from the Decision Tree node by opening the Explorer window inside EM, click the checkbox for Show Project Data, then navigate to your workspace folder (Emws15 e.g.). All the data sets with the TreeZ prefix are the ones created by the Decision Tree node.

WendyCzika · ‎02-12-2016

Make sure you use 2 periods before tree38: inmodel=&em_lib..tree38_emtree

WendyCzika · ‎02-02-2016

For this issue and the other one that you had with the Survival node, you are probably best off going through SAS Tech Support for assistance: http://support.sas.com/techsup/contact/ Thank you! Wendy

WendyCzika · ‎02-01-2016

Yes as Reeza mentions, I don't think PMML score code is the way to get there. See the Reference Help on the Score Code Export node. This can take the DATA step score code that is produced for a flow and create the files needed to deploy in a database.

WendyCzika · ‎01-13-2016

When you are creating your data source, be sure to select Advanced in the Metadata Advisor as shown below, and it should set the level for your binary variables to BINARY automatically. The Basic Advisor assigns levels based on the type of variable only.

Online Status	Offline
Date Last Visited	a week ago