How Do I Get Started with SAS Visual Data Mining and Machine Learning? Slides and Q&A

[#1]

Did you miss the Ask the Expert session on How Do I Get Started with SAS Visual Data Mining and Machine Learning? Not to worry, you can catch it on-demand at your leisure.

 

Watch the webinar

This session provides information on how SAS Visual Data Mining and Machine Learning is designed for the data scientist, statistician and advanced business analyst.

Whether you want to program or point and click, SAS Visual Data Mining and Machine Learning provides innovative algorithms and fast, in-memory processing.

This session covers its capabilities with an accompanying demonstration showing the components of SAS Visual Data Mining and Machine Learning.

You will learn how to:

• Interactively program in a web-based development environment.
• Use intelligent automation, including Automatic Feature Engineering node for automatically cleansing, transforming and selecting features for models.
• Use natural language generation.
• Incorporate embedded support for Python and R languages.
• Use highly scalable, distributed in-memory analytical processing.

Here are some highlighted questions from the Q&A segment held at the end of the session for ease of reference.

Does SAS Visual Statistics come with SAS Visual Data Mining and Machine Learning? 

As part of the VIYA platform when you license SAS Visual Data Mining and Machine Learning you also have SAS Visual Analytics and SAS Visual Statistics. 

Can I add SAS Visual Data Mining and Machine learning to my current SAS install? 

SAS Visual Data Mining and Machine Learning is a part of the new VIYA platform. SAS VIYA and SAS 9.4 can be integrated and interact with each other and they are a separate installation and configuration. 

Is it possible to download this data? 

This data is available with SAS Enterprise Miner as a sample dataset. You can find it in the SAMPSIO library. 

Is autotuning available in the Visual Interface? 

In the current version of SAS VDMML 8.5 autotuning is available through both the visual and programming in SAS VIYA or Jupyter Notebooks or through the task in SAS Studio. 

Is there integration with R? 

Yes, Python, Java, R and Lua are supported. You can call SAS VDMML using your preferred language in our world-class, governed environment. And using REST APIs, you can add the power of SAS Analytics to your custom applications. 

Is this part of SAS VIYA? 

Yes. SAS Visual Data Mining and Machine Learning (VDMML) is one of the SAS Viya products. 

How do you get to SAS Drive? 

Your SAS Administrator can provide you with the URL or Web address to give you access to SASDrive. 

What is the limitation of running machine learning models in Visual Analytics like this, in comparison to running Model Studio. I understand Model Studio is the full-brown ML platform? 

The visual interface shown give you access to Machine Learning techniques through a point and click visual environment. Many of the options are available in this interface. Model Studio or pipelines has additional options including the ability to build our entire Machine Learning process flows through point and click and the integration of open source and Natural Language generation. Programming will provide more options including both Procedures and CAS Actions. 

What are the measures to compare the models? 

You have several options with VDMML. You can compare with Lift, KS, and Misclassification, just to list a few. 

Can we build models without VDMML, just using visual statistics and using python? 

Absolutely, You can build models with point and click in Visual Statistics or you could open up Jupyter Notebooks and model with python. 

What do I use to manage clinical chemistry patient data? Patient had 3 different data results from 1) normal healthcare site; 2) community general hospital site and 3) area ER site that are not consistent. 

SAS Viya has several tools for helping you manage data. Within the interface showed you there are two menu options under the Hamburger or Action menu, Manage Data and Prepare Data. Manage Data allows you to bring data (including SAS, database data, or local files like .csv) into the CAS environment. Prepare Data allows you to join or append multiple data sets, transform columns, use custom code, filter, partition and more.  

After using the template that automatically estimate hyper parameters, what the indication for the need for further tuning? 

I recommend that you examine diagnostic statistics to determine how well the model is performing. Misclassification rate or ROC for example. 

How does SAS Enterprise Miner fit into the picture? 

Enterprise Miner(EM) is a client/server available with our SAS 9 platform.  If you have a newer version of Enterprise Miner you can integrate with Viya Visual Data Mining and Machine Learning if its licensed. You can run Viya models from within EM and you can run batch code from EM and score code from EM in Viya 

I have not used Visual Analytics but I have been using Enterprise Miner. What advantages Visual Analytics provide over EM on machine learning? 

Visual Data Mining and Machine Learning is provided on the Viya platform which mean it is cloud based and the data is loaded in memory. It is utilizing the newest architecture and technology. VDMML allows easier integration with open source. Also Visual Data Mining and Machine learning, provide more machine learning algorithms and options for those algorithms along with the ability to hypertune the model.  When using tools in Viya you have one interface (the browser) to use to access and modify your data and move between your analytic tasks from modeling to visualizing results. 

We are using SAS Enterprise Guide, what shall we do if we want to purchase Viya? Is it a different product? 

SAS Enterprise Guide can be an interface to the SAS Viya environment. You would need to license, install and configure Viya to work with SAS Enterprise Guide. Please contact your SAS representative to inquire about purchasing a license for SAS Viya. We can make sure to pass your name to the Sales representative if you'd like. 

What is ML with Explanation? 

Explanation provides natural language to explain the results of your analysis. An example showed were the insights shown when doing pipeline comparison. 

Is this SAS Visual Dating Mining and Machine Learning useful for data cleaning? 

VDMML is part of the Viya platform which has components for data cleaning. VDMML itself can do things like imputation and feature transformation, but data management can be done with SAS Studio and Prepare Data. 

How about build models module? I checked our environment and we don’t have 'build models' whereas we do have 'manage models'. 

It appears you don't have VDMML licensed in your environment. 

Is it possible to get access to SAS visual analytics to learn basic machine learning techniques for about two weeks? 

Use this link to start a free trial. Orange button at the top right: https://www.sas.com/en_us/software/visual-data-mining-machine-learning.html 

Documentation for Bayesian Networks? 

Link to documentation for Bayesian Networks 

Can VIYA do ensembles? 

Yes, Visual Data Mining and Machine Learning can ensemble your models. Ensemble is one of the nodes available in pipelines. 

Is there Python integration in SAS visual analytics? Yes. Viya in general has Python integration. You can integrate with Python in Pipelines, SWAT package in Jupyter Notebooks, and Proc FCMP in SAS Studio. 

Does python embedded in SAS Viya or does it have to be install where the SAS engine is? 

Python needs to be installed with SAS. 

Do you have documentation on VDMML? 

https://support.sas.com/en/software/visual-data-mining-and-machine-learning-support.html