SAP Machine Learning

Machine Learning Features in SAP Products

What is Machine Learning?

Machine learning (ML) is a subset of artificial intelligence (AI) algorithms. The differentiating aspect of these algorithms is that they can learn from the input data and modify the model based on changes in that data. It is this “learning” aspect that makes these algorithms powerful.

Machine Learning Applications in SAP Portfolio

SAP applications leverage ML algorithms extensively to embed innovative capabilities within their solutions, help end-users perform advanced analytics with minimal technical proficiency, and allow data scientists and ML engineers to build advanced models and solutions. Below are some examples:

• SAP HANA

SAP HANA has been designed to be easily leveraged as a scalable ML platform. A powerful built-in tool is the predictive analytics library (PAL). A component of the application function library in HANA, PAL includes several algorithms to enable the most frequently used predictive analytics use cases. For advanced users who want to explore advanced algorithms like deep learning, extended machine library (EML) in HANA allows such users to leverage TensorFlow to build deep learning algorithms.

• SAP Data Intelligence

SAP data intelligence has a rich ML content library. This library, which has an ML scenario manager and ML operations cockpit, allows engineers and data scientists to collaborate and build ML models.

• SAP Analytics Cloud Smart Predict

Like most best-of-breed analytics tools, SAP Analytics Cloud provides users the ability to leverage advanced ML algorithms. While ML algorithms have many applications, predictive analytics remains a key one.

Key Considerations for SAPinsiders

Machine Learning Features in SAP Products

What is Machine Learning?

Machine learning (ML) is a subset of artificial intelligence (AI) algorithms. The differentiating aspect of these algorithms is that they can learn from the input data and modify the model based on changes in that data. It is this “learning” aspect that makes these algorithms powerful.

Machine Learning Applications in SAP Portfolio

SAP applications leverage ML algorithms extensively to embed innovative capabilities within their solutions, help end-users perform advanced analytics with minimal technical proficiency, and allow data scientists and ML engineers to build advanced models and solutions. Below are some examples:

• SAP HANA

SAP HANA has been designed to be easily leveraged as a scalable ML platform. A powerful built-in tool is the predictive analytics library (PAL). A component of the application function library in HANA, PAL includes several algorithms to enable the most frequently used predictive analytics use cases. For advanced users who want to explore advanced algorithms like deep learning, extended machine library (EML) in HANA allows such users to leverage TensorFlow to build deep learning algorithms.

• SAP Data Intelligence

SAP data intelligence has a rich ML content library. This library, which has an ML scenario manager and ML operations cockpit, allows engineers and data scientists to collaborate and build ML models.

• SAP Analytics Cloud Smart Predict

Like most best-of-breed analytics tools, SAP Analytics Cloud provides users the ability to leverage advanced ML algorithms. While ML algorithms have many applications, predictive analytics remains a key one.

Key Considerations for SAPinsiders

• Develop a fundamental understanding of algorithms: Explore what specific algorithms are available and understand where they can be leveraged. This will help you get optimal value from these tools. As an example, you should be aware that you can use clustering algorithms for customer segmentation. Here is an example of a good overview of critical algorithms used in SAP applications.
• Understand the limitations of underlying data infrastructure: Understanding aspects of the underlying database is also critical. This helps you build pragmatic models. As an example, HANA has a 2 billion rows limitation, and hence you may have to partition tables for data sets larger than that. This impacts your model development as well.
• Understand the limitations of tools available: Some PAL algorithms have limits on the number of parameters. This means you will have to pay more attention to feature selection or feature engineering while building models with these algorithms. You can find several examples of these limitations on the SAP help portal and SAP blogs.