SAP HANA APL Hands-On — AutoClassifier, AutoRegressor, AutoTimeSeries and Drift Detection
As of 2026-09-25
Hands-on module on SAP HANA's Automated Predictive Library (APL), the automation-first sibling of PAL covered in M338. Explains what APL is (SAP's own definition: exposing the Automated Analytics engine's data mining capabilities through SQL), its model types (AutoClassifier, AutoRegressor, AutoTimeSeries, AutoUnsupervisedClustering, supervised clustering, DriftDetector, and direct gradient-boosting classes), what automation removes from a data scientist's to-do list (feature encoding, variable selection, hyperparameter tuning), how to call it (SQL stored procedures or the hana-ml Python client) and read its build_report() output, drift detection as a concrete retraining trigger, and the decision framework for PAL vs APL vs SAP-RPT including an explainability requirement. Three exercises compare an APL result against the equivalent PAL result from M338 on the same data.
What you will learn
- Explain what APL automates that PAL leaves manual — feature encoding, variable selection and hyperparameter tuning — and why that trades control for speed
- Name APL's model types (AutoClassifier, AutoRegressor, AutoTimeSeries, AutoUnsupervisedClustering, supervised clustering, DriftDetector) and pick the right one for a business question
- Fit an AutoClassifier or AutoRegressor with hana-ml and read its build_report() output for model quality and variable contribution
- Use DriftDetector to compare a training snapshot against a recent extract and state what should trigger retraining
- Compare an APL result against the equivalent PAL result from M338 on the same data, and explain the trade-off in plain language
- Choose between PAL, APL and SAP-RPT for a given business question, including cases where a written explanation requirement changes the answer
Module overview
Who this is for. You have worked through M338 (SAP HANA PAL Hands-On) or already know PAL well enough to ask, reasonably, "why would I not always just tune PAL myself." This module answers that question hands-on with the Automated Predictive Library (APL), SAP HANA's automation-first machine-learning engine. M340 goes deeper into the Python client, hana-ml, that wraps both PAL and APL; M341 shows how a trained model — PAL's or APL's — is industrialised in S/4HANA through Intelligent Scenario Lifecycle Management (ISLM). Work through this module with a HANA Cloud instance if you can reach one; every exercise also has a paper-only path.
Prerequisites
- M338 (SAP HANA PAL Hands-On) or equivalent familiarity with PAL's classification, clustering and association capabilities
- Comfort reading SQL and basic Python; the exercises reuse the same labelled table pattern as M338
- Optional for the hands-on variants of E1-E3: access to a SAP HANA Cloud trial or client sandbox with APL enabled, and the hana-ml package installed
Outcomes
- Explain to a client what APL automates and what it gives up compared with PAL, in language a non-technical sponsor can follow.
- Fit an APL model end to end and read its build_report() output for model quality and variable contribution.
- Use DriftDetector to decide when a deployed model needs retraining, and name who owns that decision.
- Recommend PAL, APL or SAP-RPT for a given SAP prediction use case, including when an explainability requirement changes the answer.
Full module available to members. The full module adds: the decision framework · the end-to-end scenario walkthrough · the KPI scorecard · the anti-patterns · the code blocks · the knowledge check · the diagrams.