Professional Documents
Culture Documents
Advanced Analytics For The Non-Data Scientist Slides
Advanced Analytics For The Non-Data Scientist Slides
Advanced Analytics For The Non-Data Scientist Slides
A human using
a computer
A computer
providing info
to a human
“Advanced Analytics” Is the Next Step in Business Intelligence
Representative
User Base
Specialization
Chasm
PM
LOB DM
ETL
PA
ETL
Reporting Discovery
Cleaned Data BI
EIM Analysis Dashboards
Raw Data
Specialization
PM
LOB DM
ETL Automated
Predictive PA
“Invisible
Analytics” in BI
Reporting Discovery
BI
Cleaned Data BI
EIM Analysis Dashboards
Raw Data
Powered by
Predictive
Analytics
Powered by
Predictive Analytics
Reduce costs:
A major Canadian bank:
– Increased campaign response rates by 600%
– Cut acquisition cost by 50%
– Boosted ROI by 100%2
A European telecom reduced customer churn from 20% to 5% using predictive analysis1
Airlines better estimate the number of passengers who won’t show up for a flight2
Save Lives:
Health care: finding emerging symptoms for Ebola before the pattern is obvious to the naked eye
Route optimization to reduce response times: how much is 5 minutes worth to a dying patient?
1 The Economist, The Data Deluge, “Data, data everywhere”, February 27,2010, pages 3-5
2 Wayne W. Eckerson, Predictive Analytics: Extending the Value of Your Data Warehousing Investment, TDWI Best Practices Report, 2007, page 6
* “Advanced Analytics” and “Predictive Analytics” will be used relatively interchangeably in this presentation
High level:
Using algorithmic analysis to recognize data relationships that influence likely outcomes and identify potential
risks and opportunities before they occur to make better decisions in the future.
– Approximate the relationship between variables and their outputs and represent it as an algorithm (set of rules)
– Use the algorithm against future data to predict the response with the least amount of error
In the BI context:
PA is another toolset that helps us uncover patterns and relationships algorithmically that can be used against
future data sets instead of relying solely on visual representations that require the analyst to infer the future
Trivial Example:
Analysis of an entire year’s worth of store sales data reveals the following patterns:
– Soda is bought 40% of the time when potato chips are purchased
– Purchases of umbrellas increase 125% during the months of March and April
– When baby formula is purchased by a male: 54% of the time, so is at least one case of beer
+
© 2015 SAP SE or an SAP affiliate company. All rights reserved. 16
Predictive 101
(Advanced analytics for the rest of us)
A predictive model algorithmically represents the desired target within the dataset. This model
can then be applied to other (or future) datasets to identify elements that should be targeted.
Predictive Models:
“This last transaction could be fraudulent”
Descriptive Models:
“Male purchasers of baby formula are 56% likely to also buy a case of beer, but female purchasers are likely to
have a higher number of items and a larger total bill”
Decision Models:
“Your credit score is pretty low and you have few assets, giving you a mortgage would be pretty risky …”
SAP Predictive Analytics Automated Algorithms are adaptive and are self-training
Model can be validated by applying against the “testing” set of historical data to determine “Predictive Power”
System iterates continuously until predictive power is high enough
Constant validation to determine when model needs to be modified (Model Manager)
What they What they put What they When they How they How much What else they What do they
searched for in their cart actually bought bought bought they paid bought buy regularly?
Good News! There’s no “100% right” answer – There’s just different levels of “right”…
NEW Data
(Current Customers)
Targeted List
Name Gender Age Marital Recent Activity C-Sat Renewed Before Churn
Real-Time Scoring
NEW Data
(Single Customer)
Model
Special Offer:
Hancock, John M 38 D Y 4.2 N Y
Renew now and save 25%!
Embedding Predictive Analytics Into BI Workflows
Model Embedded as
Stored Procedure
Customer Database
Business Users can get on-the-fly
Hancock, John M 38 D Y 4.2 N Y scoring without even knowing they
SQL are using predictive algorithms
Doe, Jane F 45 M Y 9.4 N N
You don’t need to know HOW the algorithms work, but you do need to know WHAT they are for …*
Advanced Analysis
… … … …
… … … …
… … … …
* Actually, it does help to know how they work, but it is not absolutely required to get started
30
SAP Solutions for the Entire Spectrum of Users
No Low High
Business Data
Business Users & LOB
Analysts Scientist
Relational Databases Big Data Sources CRM/ERP Stores Applications Cloud Services
Transactions
Cloud / On-Premise
No part of this publication may be reproduced or transmitted in any form or for any purpose without the express permission of SAP SE or an
SAP affiliate company.
SAP and other SAP products and services mentioned herein as well as their respective logos are trademarks or registered trademarks of SAP SE
(or an SAP affiliate company) in Germany and other countries. Please see http://global12.sap.com/corporate-en/legal/copyright/index.epx for additional trademark
information and notices.
Some software products marketed by SAP SE and its distributors contain proprietary software components of other software vendors.
These materials are provided by SAP SE or an SAP affiliate company for informational purposes only, without representation or warranty of any kind, and SAP SE or its
affiliated companies shall not be liable for errors or omissions with respect to the materials. The only warranties for SAP SE or
SAP affiliate company products and services are those that are set forth in the express warranty statements accompanying such products and services, if any. Nothing
herein should be construed as constituting an additional warranty.
In particular, SAP SE or its affiliated companies have no obligation to pursue any course of business outlined in this document or any related presentation, or to develop or
release any functionality mentioned therein. This document, or any related presentation, and SAP SE’s or its affiliated companies’ strategy and possible future
developments, products, and/or platform directions and functionality are all subject to change and may be changed by SAP SE or its affiliated companies at any time for
any reason without notice. The information in this document is not a commitment, promise, or legal obligation to deliver any material, code, or functionality. All forward-
looking statements are subject to various risks and uncertainties that could cause actual results to differ materially from expectations. Readers are cautioned not to place
undue reliance on these forward-looking statements, which speak only as of their dates, and they should not be relied upon in making purchasing decisions.