Watch this Ask the Expert session to explore how to fit, evaluate and compare machine learning models with ease in SAS Viya.
Watch the webinar
You will learn:
- What a machine learning model is, including different types.
- How to assess and compare machine learning models.
- How to fit multiple machine learning models, including open source models, using SAS Viya.
The questions from the Q&A segment held at the end of the webinar are listed below and the slides from the webinar are attached.
Q&A
If I, at some point, notice that there are some outliers/faulty data rows, can I remove/filter them out in SAS, or do I have to do it elsewhere and load a new dataset to SAS?
You can certainly do that within SAS. It depends on your platform. For example, in Model Studio right go to the pipeline. In the pipeline nodes, some of the nodes under data mining and preprocessing, such as anomaly detection, are designed to find outliers, and you can have them removed. There is also filtering, which does not have to be limited to outliers. For instance, you might have four different job types in your data, but you don't want the model to apply to the last category —so you can use a filtering node to exclude those observations. In short, you can accomplish all of this within SAS, and there are multiple ways to do it depending on your interface.
In what ways can SAS Viya be used to ensure reliable performance when selecting the best model for a specific business problem?
I am not sure what is meant by reliable performance. There could be several different meanings for that phrase. I hope that I am correct by assuming that you mean that you want a model that will reliably predict the response.
For this situation, there are a few things that can be done. By using honest assessment (splitting the data into training and validation sets – and possibly a test set) you will receive a reasonable idea of real-world performance of the model. SAS Viya offers you options to split the data into these partitions.
Further, there are several different metrics that can be used to select the “best” model for a problem. You can choose which metric SAS Viya will use to determine the champion model. But all the other metrics are available as well. By considering multiple metrics, specifically the ones that relate to the desired usage of the model, you can ensure that you have obtained a model that is more robust and reliable to your specific business needs.
Recommended Resources
Course: Machine Learning Using SAS® Viya®
Course: The Modeling Life Cycle for Data Scientists
Course: Statistics You Need to Know for Machine Learning
Please see additional resources in the attached slide deck.
Want more tips? Be sure to subscribe to the Ask the Expert board to receive follow up Q&A, slides and recordings from other SAS Ask the Expert webinars.
