Part 8 of the blog series:
A podcast on the topic is also available here.
Let us now dive into another approach of doing machine learning with SAP S/4HANA business processes. In the earlier blogs, we discussed about how to embed machine learning directly into SAP S/4HANA applications and consume the machine learning services from SAP Cloud Platform into the SAP S/4HANA business processes. Now we will work our way through leveraging the SAP Analytics Cloud – Smart Predict and Smart Assist services to do machine learning on the SAP S/4HANA business processes.
Typically data scientists need to work with business users or business experts to understand the requirements in detail to design and build the predictive models. Even if all these things happen as planned, the end result might still vary due to gaps in the business knowledge and expertise needed by the data scientists. While the business users and business experts can very well understand the needs to design a model, they are not equipped like a data scientist to build these models. Hence with the introduction of SAP Analytics Cloud, by leveraging the Smart Predict and Smart Assist services, the business users can design and build predictive models that work seamlessly to solve the business problems. These SAP Analytics Cloud predictive services could help the business users do data science models even though they are not comfortable with statistics and algorithms. This helps the data scientists be freed from these mundane tasks of creating predictive models and focus their efforts on building more important machine learning functionality. With SAP Analytics Cloud, the focus has now turned on working with the business questions and not algorithms – changing the way of how you deal with creating powerful solutions which helps speed up the prediction and recommendation process.
In this approach, we do explorative predictive analytics by leveraging the side-by-side mode of taking a CDS view (core data services) from SAP S/4HANA business process, build a custom CDS view and then connect to SAP Analytics Cloud for doing predictive modeling and analysis. Depending on the use case requirement and functionality, you can do a live connection from SAP Analytics Cloud to SAP S/4HANA core data services (CDS) views or do a data acquisition approach to build additional virtual data models (VDM) in SAP Analytics Cloud. You can refer to this blog series which talks in detail about how to make the configuration connections between SAP S/4HANA and SAP Analytics Cloud, then create predictive models using the Smart Predict functionality in SAP Analytics Cloud with visualization of these predictions.
Smart Predict addresses most business problems by leveraging internally the algorithms such as Time series forecasting, Classification analysis, Regression analysis, Smart grouping etc. Generally the prediction results provided while using the Smart Predict or Smart Assist services are not permanently stored into the business processes but saved as templates or dashboards. These dashboards can help the business users analyze the prediction results and leveraged appropriately. If it is deemed necessary, these prediction results can be integrated back into the SAP S/4HANA business processes by publishing the predictive models into SAP S/4HANA using the PAi framework or implement the functionality as an embedded ML or side-by-side ML as explained in the earlier blogs. The new ISLM framework has not yet been integrated into SAP Analytics Cloud and will update once available. I will publish a blog later on the roadmap and upcoming functionality.
The following gives a quick overview of the process involved in the creation of the predictive models using SAP Analytics Cloud – Smart Predict and Smart Assist services. We are briefly explaining here in this context for the sake of completion but would not go into the details since it can be found in this blog with detailed instructions.
Step1: Menu path to create a Predictive Scenario in SAP Analytics Cloud
Step2: Selecting the type of Predictive Scenario
Step3: Create a time series Predictive Scenario
In the next blog let us discuss how you can extend an already created ML scenario.
Here are some quick links to the blogs in this series to give you a complete understanding of how Predictive Intelligence is infused into SAP S/4HANA.
- Resources and journey to machine learning with SAP S/4HANA
- Part 1 – Leveraging Predictive Intelligence with SAP S/4HANA
- Part 2 – Architecture and deep-dive of the different approaches around Predictive Intelligence
- Part 3 – Process flow leveraging Machine Learning and Predictive Analytics
- Part 4 – Scope and functionality in the context of an end-to-end process leveraging ML
- Part 5 – Activating machine learning functionality for SAP S/4HANA
- Part 6 – Building ML into the digital core of SAP S/4HANA (Embedded ML)
- Part 7 – Enhancing the digital core with ML Services (Side-by-Side ML)
- Part 8 – Extending the digital core by leveraging ML with SAP Analytics Cloud (this blog)
- Part 9 – ML Extensions to SAP S/4HANA processes
- Blog series – ISLM for machine learning with S/4
- Introducing the book – Implementing Machine Learning with SAP S/4HANA
Happy predicting the future!!