5 things you should know about SAS® Visual Forecasting


SAS Visual Forecasting is a powerful and flexible tool that can cover all the forecasting needs of an organisation. Users can automatically produce large-scale time series analyses and hierarchical forecasts – without human involvement. Having both a friendly user interface and a programmatic interface allows everyone to take part in the forecasting process.

But what sets it apart from the competition? The list is long, but let’s focus on 5 key capabilities that users may not be familiar with.

1. Open source scalability

SAS Visual Forecasting can distribute open source code (Python and R) to run in parallel on the same nodes where SAS® Viya® is installed. In that way, you can easily scale up open source forecasting processes to millions of series. And move away from ungoverned, inconsistent and error-prone processes that run on users’ local desktops. For more information on parallelization, make sure to check the paper here. And if you are looking for a practical example on how to parallelise the FB-Prophet open source algorithm using SAS Viya, make sure to check the blog post here.

2. Time-series segmentation

SAS Visual Forecasting comes with a segmentation capability based on the time-series patterns it detects in the user’s data. As a result, users can apply different modelling strategies to different segments. Not every series requires manual intervention from modellers, especially when we are dealing with hundreds of thousands of series. Arima and exponential smoothing models, which the system automatically generates and optimizes, provide surprisingly accurate results when we have enough data history and don’t deal with very complex patterns.

By automatically segmenting time-series based on their patterns (volume, volatility, seasonality, intermittency, etc.), users can develop modelling pipelines for each one separately and focus on developing sophisticated forecasting modelling strategies on the series that require the most attention. Have a look here for detailed information on automatic segmentation. Or if you want to bring your own segments, have a look here to learn more about the external segmentation pipeline.

3. Deep learning

SAS Visual Forecasting lets you run recurrent neural networks (RNNs), long short-term memory (LSTMs) and gated recurrent unit (GRUs) in a simple way. This is done using the new "TNF" package. SAS automatically structures the data in the right way, saving you significant time compared to manually applying feature engineering techniques.

For a closer look, check out this resource, which includes the necessary theory to understand what’s happening behind the scenes and also some good examples to get you started. If you want to incorporate the above deep learning techniques in your SAS pipelines and compare and select the most accurate results with other forecasting algorithms at a series level, then have a look at this blog post, which takes you through the whole process.

4. Hybrid modelling techniques

SAS Visual Forecasting includes proprietary nodes that use neural networks in combination with traditional time-series techniques to provide the most accurate results. Feature engineering and data aggregation are also taking place automatically, saving massive amounts of time for users.

Something to remember when using these techniques is that the model is trained on all data. So you don’t get different models per series as, for example, you would with RNNs, Arima, exponential smoothing, etc. These techniques work great when we want to incorporate many variables in our forecasts or have to deal with complex patterns in our data. Have a look at this paper that takes you through all the available techniques in more detail.

5. Build your own nodes

SAS Visual Forecasting gives flexibility to users to create their own custom nodes and share them as plug-and-play solutions with other data scientists and forecasters around the business. The most efficient way to do this is by using "The Exchange," which is a space where SAS Viya users can exchange assets they create in a simple and robust way. Alternatively, you can also upload custom nodes as zip files directly to SAS Viya.

For a step-by-step guide, have a look at this resource, which demonstrates the process of developing a manual gradient boosting node that you can load and use directly in any forecasting pipeline.

That’s all for now. Stay tuned as new capabilities are released monthly. This list is just going to get bigger!


About Author

Spiros Potamitis

Spiros Potamitis is a data scientist at SAS, specialising in the development and implementation of advanced analytics solutions across different industries. Having acquired an MSc in Computer Engineering and one in Information Management Spiros provides subject matter expertise in the areas of Forecasting, Machine Learning and AI. Prior of joining SAS, Spiros has worked and led advanced analytics teams in various sectors such as Credit Risk, Customer Insights and CRM.

Leave A Reply

Back to Top