Technology Blogs by SAP
Learn how to extend and personalize SAP applications. Follow the SAP technology blog for insights into SAP BTP, ABAP, SAP Analytics Cloud, SAP HANA, and more.
Showing results for 
Search instead for 
Did you mean: 
Artificial Intelligence (AI) and machine learning are the hottest trends in IT. Organizations are adopting these technologies to optimize processes and to process information. SAP HANA Cloud offers Machine Learning capabilities natively, which you can easily enable in a productive environment.

With the embedded machine learning in SAP HANA Cloud, it is possible to build comprehensive multi-model, in-database applications. You can develop embedded machine learning scenarios using trending ML (Machine Learning) algorithms processed with SAP HANA in-memory performance. It’s also possible to combine and enrich machine learning applications with spatial and graph processing in SAP HANA Cloud. On top of all of that, SAP HANA Cloud also supplies R and Python machine learning client interfaces for Data Scientists.

SAP HANA Cloud embeds several machine learning libraries, designed and optimized for massive parallel in-memory processing. There are two major components available within SAP HANA Cloud which can be used for machine learning:

  • PAL – Predictive Analysis Library

  • APL – Automated Predictive Library

Predictive Analysis Library (PAL)

PAL is a native (built-in) function library (AFL) for SAP HANA with many specialized algorithms. APL bundles the algorithms of SAP Predictive Analytics and automates most of the steps in applying machine learning algorithms, like data preprocessing or variable selection. Both libraries provide algorithms for classic machine learning use cases like classification, regression, time series forecasting, cluster analysis, and more.

The over 90 PAL algorithms include trending algorithms like Random and Gradient Boosting decision trees including automated cross validation and hyper model parameter selection for fastest building of best models. The PAL algorithms support high-performance parallel mass prediction as well as real-time transactional speed predictions. Furthermore, segmented machine learning models can be composed, like segmented forecasting models by store for example

New in SAP HANA Cloud PAL includes several algorithms that support   continuous learning and update of the e.g. regression model, thus enabling dynamic and real-time machine learning, where the predictions models get instantly adapted to the changing conditions in current data.

Learn more about Machine Learning libraries (APL and PAL) in SAP HANA Cloud.

Automated Predictive Library (APL)

The APL is a highly automated ML framework also including Gradient Boosting based classification and regression algorithms. APL exposes complex machine learning functions to non- expert users via simple procedure interfaces for developers to Create, Train, Apply, Deploy and Query predictive models, implicitly hiding and automating the expert’s tasks for using machine learning algorithms.

Using PAL and APL with SAP HANA Cloud

To enable the machine learning capabilities in SAP HANA Cloud, you need to enable the ScriptServer within your SAP HANA Cloud database. You can do that either when creating a new instance or by editing an existing instance. This is easily done with the self-service capabilities within SAP Cloud Platform. However, the ScriptServer requires that the SAP HANA Cloud database includes at least 3vCPUs (or 45GB of memory) to be enabled. The instance size can also easily be scaled-up by using the self-service capabilities.

Both PAL and APL can be called through a SQL interface or via the Python and R machine learning client’s interfaces, allowing Data Scientist to easily tap into building SAP HANA Cloud embedded machine learning scenarios directly from their expert environment.

Get started with this hands-on tutorial.

Find more resources on Machine Learning with SAP HANA.