Send the link below via email or IMCopy
Present to your audienceStart remote presentation
- Invited audience members will follow you as you navigate and present
- People invited to a presentation do not need a Prezi account
- This link expires 10 minutes after you close the presentation
- A maximum of 30 users can follow your presentation
- Learn more about this feature in our knowledge base article
Do you really want to delete this prezi?
Neither you, nor the coeditors you shared it with will be able to recover it again.
Make your likes visible on Facebook?
You can change this under Settings & Account at any time.
SAP Business Analytics Symposium 2013
Transcript of SAP Business Analytics Symposium 2013
we know Q. How can we create information we never collected?
A. *Sometimes* we can infer it Facilities all over
the world 4 millions plus materials in the system Over 300 MRP
Controllers Significant percentage of unknown values
Fed from engineering systems or manual override
Default values unrealistic usually 999 Approach Data Audit Algorithm Selection Execution and Results Load Data Join Tables Configure Predictive Models Execute & Analyze Initially nothing is known about the data. We needed to:
Understand the data
Prepare it for analysis
Ingest into an analytical platform
Build and test predictive models
Disseminate back into the production systems
The SAP Predictive Analysis software integrates into the clients existing infrastructure and is the logical choice. We undertook an audit of the data. We needed to understand:
What values are factors?
What values are numeric?
Which values are complete?
SAP PA is designed to integrate well with EDW ... but we didn't have one ... Depends on the question
Working towards material plant lead times; a numeric
SAP PA supported algorithm library Three data sets:
Measure the error SAP systems generate huge amounts of exhaust data
If no one's using it, then you're just wasting drive space
This data can be recycled given the tools and people Tomorrow Today All about the exhaust ... Peter Owlett @PeterOwlett Accuracy Split
Re-bin the values
Segmentation and re-prediction By populating the material lead time field, the client can more effectively leverage their SAP ECC System Better planning data gives more accurate plans Pattern Detection HANA Integration The models we built can be reused and refined over time to maintain an ongoing capability SAP PA support PMML and integrates with HANA, allow models to be stored, managed and executed from HANA A set of statistical and machine learning approaches that take a set of labeled data observation and produce an reusable data model object. A way to estimate something that is currently unknown Lots going on right now
Methods We're going to look at using predictive analytics in a materials management context 56
458 <Screenshot removed> <Screenshot removed>