IBM Analytics
IBM SPSS Predictive Analytics
IBM Analytics
Business landscape changes: Digital Disruptors
World’s Largest Accommodations Company... Owns No Real Estate
World’s Largest Taxi Company...
CEOs Expect More Digital Interaction 2015
82%
+37
%
2013
60%
Owns No Vehicles
CEO’s Expect More Competition From Other Industries 2015
2013
60%
40%
+50
%
World’s Largest Retailer...
World’s Largest Media Company...
Carries No Inventory
Creates No Content
3
Insights from IBM’s Global C-suite Study — The CEO Perspective ibm.com/csuitestudy
IBM Analytics
Analytics-driven organizations reap rewards 66 50
% %
Torchbearer CEOs
32%
Market Follower CEOs
Insights from IBM’s Global C-suite Study – The CEO Perspective (2015) ibm.com/csuitestudy
Use Predictive Analytics
Front Runners Outperform On Business Outcomes (69%)
Competitive Advantage (53%)
By Using Data and Analytics Source: IBM Institute for Business Value (IBV) (2013)
4
Revenues (60%)
IBM Analytics
Predictive analytics helps you uncover the value in big data
See patterns and trends, connect data to effective action and apply insights throughout your business.
5 ibm.biz/predictive
IBM Analytics
It provides answers to critical questions … How can I find my best customers? Locate stores or facilities? See how competitors affect my market?
How do I predict behaviors and preferences so I can reduce customer churn, prevent fraud, improve processes, or grow revenue?
How can I act in real time or ahead of a potential issue or event, instead of after it happens?
6 ibm.biz/predictive
IBM Analytics
…and helps guide your organization with data-driven actions Seize new opportunities Manage risk Implement more effective strategies
7 ibm.biz/predictive
IBM Analytics
Nearly half of all organizations apply predictive analytics in areas critical to revenue and profitability
48% of organizations surveyed apply predictive analytics to their marketing activities.
44% apply predictive analytics to improve operations. --Next-generation predictive analytics: Using forward-looking insights to gain competitive advantage, Ventana Research, 2015
8 ibm.biz/predictive
IBM Analytics 95% of
survey say datapredictive and Nearly half of allrespondents organizations apply analyticsinare necessary to revenue keep them onprofitability par with analytics areas critical to and or ahead of their competitors. - Analytics: The upside of disruption, IBM Institute for Business Value, 2015
9 ibm.biz/predictive ibm.biz/predictive
IBM Analytics
Advanced Analytics are Pervasive Each Application is a Blueprint for the Next…and the Next…and the Next.. Breadth of analytic use as reported by respondents: In 2014,
In 2015,
and
10%
71%
33%
of organizations were using advanced analytics in three or more functional areas of their business
of organizations are using advanced analytics in three or more functional areas of their business
of organizations are using advanced analytics in six or more functional areas of their business
Advanced analytics are defined as the extensive use of predictive, prescriptive or cognitive analytics within a business function Source: Analytics: The upside of disruption. IBM Institute for Business Value 2015 Analytics research study. © 2015 IBM Institute for Business Value.
10
IBM Analytics
Advanced Analytics is For Everyone
Advanced analytics is the analysis of all kinds of data using sophisticated quantitative methods (for example, statistics, descriptive and predictive data mining, simulation and optimization) to produce insights that traditional approaches to business intelligence (BI) — such as query and reporting — are unlikely to discover. An advanced analytics platform provides a full suite of tools for use by knowledgeable users, traditionally data scientists. However, they are increasingly being directed at business analysts and "citizen data scientists," to enable them to perform a variety of analyses on different types of data. 11
Gartner, Advanced Analytics MQ 2015
IBM Analytics
IBM Advanced Analytics Customer analytics
Operational analytics
Threat & Fraud analytics
Acquire
Manage
Monitor
Grow
Maintain
Detect
Retain
Maximize
Prevent
IBM® (SPSS®) Predictive Analytics
Statistics
12
Modeler
Analytic Server
IBM® Prescriptive Analytics
CPLEX Studio
Decision Optimization Center (DOC)
DOCloud
IBM Analytics
IBM SPSS – The Most Comprehensive Predictive Analytics Platform
ALL Data
All Decisions
(Structured, Unstructured, Streaming)
(People, Systems, Strategic, Operational, Real-Time)
Data Preparation
13
Analytics at Scale
Insight to Action
• Data models
• Predictive models
• Real-Time Scoring
• Data connectors
• Machine Learning
• Optimized Decisions
• Data Wrangling
• Statistical Analysis
• APIs & services
• Decision Optimization
• Dashboards / Interactive apps
IBM Analytics
Organizations using predictive analytics see results
Enjoyed a 75% higher click through rate and a 73% higher sales lift than companies that did not use predictive analytics1
The ROI of business analytics solutions that incorporate predictive analytics is about
250% 2
14
1. Divide & Conquer: Using Predictive Analytics to Segment, Target & Optimize Marketing (pg. 1), Aberdeen, February 2012. 2Source: IDC, The Business Value of Predictive Analytics, June 2011
Predictive Analytics in Action
FleetRisk Advisors
U.S. Insurer
Helps trucking operators prevent accidents by building stronger & faster risk prediction models 159% boost in the average monthly rate of customers who return to shop compared to the previous year 20% reduction in minor accidents 30% increase in driver retention rates
Analyzes and links claims and medical data, to prevent fraud and fast-track legitimate claims USD 22 Millon ROI anticipated, and 100% payback anticipated from fraud reduction Over 85% accuracy for predicting independent medical failures
ASTRON
Oak Lawn Marketing
Uses streaming analytics to deliver insights from the world’s largest radio telescope 99% faster identification of data Analyzes >1 exabyte of data daily Integrates data from >3,000 dishes & antennas to form the largest & fastest radio telescope in the world
Understand customer buying patterns for targeting infomercials 159% boost in the average monthly rate of customers who return to shop compared to the previous year 400% increase in expected total revenue over a three year period 15
IBM Analytics
Predictive Analytics Use Cases
Evidence-based medicine
Human capital management
Crime prediction and prevention
Supply chain management
Process optimization
16
• Improving patient care and satisfaction • Reducing costs through optimized allocation of resources • Measuring and improving patient outcomes • Acquiring, growing and retaining employees • Helping ensure optimal staff levels • Increasing performance, efficiency and engagement • Identifying predictors of threat and fraud • Optimizing force deployment • Anticipating and visualizing crime hot spots
• Increasing visibility into virtually all areas of the supply chain • Decreasing downtime and unpredictability • Improving customer satisfaction • Improving accurate responses at the point of impact • Decreasing costs through operational efficiency • Transforming threat and fraud identification processes
IBM Analytics
IBM SPSS Predictive Portfolio Overview
IBM Analytics
IBM SPSS Predictive Analytics Predictive customer analytics
Predictive operational analytics
Predictive threat & Fraud analytics
Acquire
Manage
Monitor
Grow
Maintain
Detect
Retain
Maximize
Prevent
IBM® SPSS® Predictive Analytics
Statistics
18
Modeler
Analytic Server
Analytical Decision Management
IBM Analytics
Predictive Analytics for All
Deep Predictive Capabilities With Indatabase, R and Python Integration, and Integrated Deployment Expert User
Visual Workbench for Building Models of Any Complexity With Ability to Automate and Combine Tasks Intermediate User
Simple, Easy-to-Use, Visual, Guided Analytical Discovery, Intelligent Automation and Visual Storytelling Novice 19
IBM Analytics
Simplicity without Sacrifice
20
IBM Analytics
Open and Integrated
21
IBM Analytics
Flexible deployment
Flexible Deployment Options, Including Cloud, Supported By An Infrastructureagnostic Platform Flexible
Integrated With IBM’s High-performance Systems Built for Big Data Analytics Integrated
Embedded Into Operational, Mobile and/or Cloud-based Applications Embedded
22
IBM Analytics
IBM SPSS Statistics Quickly understand large and complex datasets using advanced statistical procedures ensuring high accuracy to drive quality decision-making
Reveal deeper
Process and deploy analytics faster with flexible deployment options 23
Programmability for advanced users that leverages common statistical programming languages in the market (Python, R)
insights and provide better confidence intervals via visualizations and geographic spatial analysis
IBM Analytics
IBM SPSS Modeler
24
Comprehensive predictive analytics platform Improve outcomes through predictive intelligence Flexible adoption and configuration -- on premises, in cloud, and everything in between Scale from personal usage, point solution(s) to enterprise deployment Providing a range of advanced analytics -decision management, text analytics, entity analytics, social network analysis and optimization.
IBM Analytics
Analytical Techniques in IBM SPSS Modeler Technique
Algorithms
Usage
Classification
Autoclassifiers, Decision Trees, Logistic, Support Vector Machines, Time Series
Predict Group Membership (e.g., Will This Employee Leave?) Or a Number (e.g., How Many Widgets Will I Sell?)
Segmentation
Autoclusters, K-Means, Anomaly Detection
Classify Data Points Into Groups That Are Internally Homogenous and Externally Heterogeneous, Identify Cases That Are Unusual
Association
Apriori, CARMA, Sequence
Find Events That Occur Together Or In a Sequence Market Basket
Geospatial
Space-Time Boxes
Ability to Improve Model Accuracy (for Any Model Type) By Including Inputs Derived From Geospatial Data Sources
Automated
Autoclassified, Autonumeric, Time Series, Clustering
Automatically Find the Right Algorithms Based On Data and Outcome to Create An Ensemble Model
Simulation
Monte Carlo
Run Different Scenarios to Identify Which Is Best From Historical Data Or Generated Data
Specialized
Text Analytics, Entity Analytics, Social Network Analysis
Improve Overall Model Accuracy
In-database
Netezza, DB2, Oracle, Microsoft
Provide User Friendly Interface On Top of Vendor Algorithms
Open Source
R/Spark/Python
Utilize Open Source Algorithms Within Modeler UI. Enhance By Easily Building Custom Dialogs (or downloading from community)
(Or Prediction)
A single analysis project may include multiple techniques 25
IBM Analytics
A marketplace-leading predictive analytics workbench
Implement an intuitive, interactive interface without the need for programming Automate modeling and data preparation capabilities Access all data—structured and unstructured— from disparate sources Apply natural language processing (NLP) to extract concepts and sentiments in text Use entity analytics to help ensure the quality of the data and results in accurate models Leverage existing investments in Cognos, PureData for Analytics (Netezza) ®, InfoSphere® and System Z® solutions
26
IBM Analytics
Analytics at Scale: Run it where the data resides IBM SPSS Analytic Server
Parallel
Optimized for Big Data environments Reduce network traffic Improved processing speed
27
In-Database
Reduce data movement SQL pushback Optimize performance with in-database adapters Increase analytic flexibility with in-database mining
IBM Analytics
Statistical Analysis & Data Mining: Feeding Predictive Analytics
Top-Down Approach A statistical approach involves forming a theory about a possible relationship converting it to a hypothesis testing that hypothesis using statistical methods It is a manual, user-driven, topdown approach to data analysis
Source: DM Review
Bottom-up Approach
Data mining involves the interrogation of the data and is performed by the data mining method rather than by the user
It is a data-driven, selforganizing, bottom-up approach to data analysis that works on very large data sets
“Statistical Modeling: The Two Cultures,” Leo Breiman, Statistical Science, 2001, Vol.16 (3), pp.199231.
Note that Both Approaches can Drive Predictive Analytics 28
IBM Analytics
IBM Advanced Analytics Customer analytics
Operational analytics
Threat & Fraud analytics
Acquire
Manage
Monitor
Grow
Maintain
Detect
Retain
Maximize
Prevent
IBM® (SPSS®) Predictive Analytics
Statistics
29
Modeler
Analytic Server
IBM® Prescriptive Analytics
CPLEX Studio
Decision Optimization Center (DOC)
DOCloud
IBM Analytics
IBM’s Decision Optimization Offering for Prescriptive Analytics
CPLEX Optimization Studio (COS) Optimization Engine
Decision Optimization Center (DOC) Development & Deployment Platform
Model complex business problems. Solve with IBM CPLEX Optimizer. Prescribe precise and logical decisions.
Build optimization solutions. Includes data analysis & visualization, scenario management, collaborative planning, and what-if analysis.
Decision Optimization on Cloud SaaS Delivery
Advanced Tools for Optimization under Uncertainty
Prescriptive analytics as a service. No install, no setup. Embed in other applications.
30
Compare multiple plans, scenarios, KPIs. Understand trade-offs e.g. costs vs robustness.Improve solution stability/robustness.
IBM Analytics
What Can Optimization Do? Optimization helps businesses make complex decisions and trade-offs about limited resources Discover previously unknown options or approaches • Automatically evaluate millions of choices Automate and streamline decisions • Compliance with business policies and regulations • Free up planners and operations managers so that they can leverage their expertise across a wider set of challenges Explore more scenarios and alternatives • Understand trade-offs and sensitivities to various changes • Gain insights into input data • View results in new ways
31
IBM Analytics
Predictive Analytics with Optimization can automate complex decisions
+
IBM SPSS
IBM Decision Optimization
=
Better decisions
All Data
Data preparation and management
Building predictive models
Predicted Data
Mathematical models + Optimizing engine
=
Optimized Decisions
IBM Decision Optimization
IBM SPSS
What are the key decisions? What are the constraints? What are the goals? 32 32
IBM Confidential
© 2012 IBM Corporation
IBM Analytics
Convergence between smart discovery and advanced analytics create new possibilities
IBM Watson Analytics
Extend
Collaborate
Transition
Smart data discovery uncovers opportunities for deeper analytics.
Meaningful analytics that a novice begins and an expert builds upon.
Make advanced analytics more consumable.
By 2018, data discovery and predictive analytics offerings will converge, with most of the leading vendors of each capability offering both. 100 Information and Analytics Predictions Through 2020, Gartner, 2015
33
IBM Analytics
SPSS and Open Source
IBM Analytics
Open Source and IBM Predictive Analytics
Embrace
First R, then Python, now Spark
35
Extend
Make coding optional
Facilitate
Make it massively useful
IBM Analytics
Extensibility with R for Modeler AND Statistics Build and score R models through the Modeler and Statistics GUIs
Use R functions for data preparation
Generate R Charts and Outputs within SPSS output management system
36
IBM Analytics
Make R usable for non-programmers
The Custom Dialog Builder for R allows you to create custom nodes in Modeler and dialogs in Statistics that run R algorithms, functions or outputs
Share nodes and dialogs with other users
End user interacts with the dialog, not code
37
IBM Analytics
& SPSS
Spark runs analytics faster on big data Complex workloads complete significantly faster in Spark compared Hadoop Map/Reduce
Spark enables users to be more productive Users are able to: 1. Build predictive models faster 2. Conduct more experiments in less time 3. Build multiple models without waiting for the system
SPSS democratizes analytics, extending benefits to users who do not want to program Access to a broader library of analytic algorithms delivers solutions to more use cases 1. In addition to SPSS algorithms that now run in Spark, Data Scientists can utilize more than 15 algorithms from Spark MLlib 2. Data Scientists can create new Modeler nodes to exploit MLlib algorithms & share them with nonprogrammer Data Scientists 3. Via shared Modeler nodes, non-programmer Data Scientists leverage Spark functionality in their own analytic workflows 38
IBM Analytics
NEW! Python for Spark Data Scientists can create extensions for novice users to exploit R, MLlib algorithms and other Python processes Spark & its machine learning library (MLlib) Other common Python libraries • e.g.: Numpy, Scipy, Scikit-learn, Pandas Abstracting code behind a GUI makes Spark usable for non-programmers
39
IBM Analytics
Programming Optional
40
IBM Analytics
https://developer.ibm.com/predictiveanalytics 41
IBM Analytics
IBM SPSS Predictive Analytics deployment
IBM Analytics
Why Is the cloud a fit for predictive analytics Capable of supporting an enterprise-scale predictive analytics environment
Eliminates the need to invest in hardware Significantly reduces time-to-value Able to support vast volumes of data
43
IBM Analytics
Ground to Cloud: Deployment Flexibility
Enterprise Analytics • Full breadth of analytical capabilities • Collaboration and enterprise-wide best practices
LOB & Personal Analytics
• Analytic tools built for business
• Digitally delivered, digitally fulfilled
Developer Tools
• Build smarter data applications -- quicker • Endless possibilities
• Customize for (virtually) any use case
• Variety of licensing and packaging options
• No installation, no configuration
• On-premises, hybrid and software-as-aservice
• Available for Windows or Mac
• Mix & match components
IBM SPSS Modeler Gold
44
IBM Cloud Marketplace
IBM Predictive Analytics on Bluemix
IBM Analytics
IBM SPSS On Digital Marketplace
LOB & Personal Analytics
• Single user perpetual/annual license
• Single user perpetual/annual license
• On-prem
• On-prem
• Base, Standard, Professional, Premium options
• Personal, Professional, Premium options
• Priced from $1K-$10K per user
• Priced from $4K-$20K per user
• Can be purchased direct from marketplace
• Can be purchased direct from marketplace
Enterprise Analytics
Developer Tools
• Enterprise solution
• Bluemix component
• IBM SPSS Modeler Gold on Cloud
• Scoring service for IBM SPSS Modeler
• Software-as-a-Service
• Subscription pricing
• $10K/month+ • Must contact sales
https://www.ibm.com/marketplace/cloud/us/enus 45
IBM Analytics
Software as a Service Model for IBM SPSS Modeler Gold:
Modeler Gold
Expand scope and breath with cloud-based statistical analysis and real-time scoring
46
Hosted environment with software, security and infrastructure managed by IBM, a recognized leader in cloud deployment
Subscription pricing with flexible terms and Tiered pricing Model
Jump start services that accelerate time to value
IBM Analytics
IBM SPSS Customer Stories
IBM Analytics
A US cable television network turns on the insights with an analytics solution that predicts success of news shows six weeks in advance
Reduces by 4x the amount of time spent processing data, from 80 – 20 percent
Triples views of video-on-demand service through data-driven marketing
Predicts success of a new show six weeks in advance of its release and adjusts marketing accordingly
Solution components • • •
IBM® Cognos® Business Intelligence V10 IBM SPSS® Modeler IBM InfoSphere® Master Data Management • IBM InfoSphere DataStage® • IBM PureData™ System for Analytics, powered by Netezza 49 • IBM Lab Services
Business challenge: This cable television network faces the challenge of managing huge volumes of information. Previously the network’s research team spent a significant amount of time processing data on spreadsheets rather than analyzing it, and based decisions on a combination of experience and instinct. The company needed a large-scale analytics solution to organize this wealth of data, make sense of it, and provide answers and actionable insights. The smarter solution: The solution combines television ratings data with information gathered minute by minute and viewer by viewer from a variety of channels and other sources to determine who’s watching and why. Then it centralizes the data and makes it available for in-depth, predictive analytics. With insights into audience preferences gained from sophisticated statistical models, the network can optimize advertising revenue and viewership like never before. Instead of paying hundreds of thousands of dollars to external vendors, the network found it could do analysis faster and more accurately in-house.
IBM Analytics
nViso – Using real-time facial imaging to deliver deep insights into customer sentiment Enables real-time analysis of complex human facial expressions
Delivers deep insights into customer sentiment in the form of structured data
Helps clients understand real-world customer behavior
The transformation: Using its facial imaging technology, nViso captures and delivers deep insights into customer reactions to its clients’ products, services and marketing communications. Thanks to a sophisticated analytics solution at the front-end, the company can help its clients uncover hidden trends to improve their offerings, better connect with customers and drive sales.
Solution components Software IBM® SoftLayer® IBM SPSS® IBM DB2®
"Using IBM software, we can empower our clients to use an entirely new source of customer data in better, smarter ways." —Tim Llewellynn, CEO and co-founder at nViso
50
IBM Analytics
Tennessee Highway Patrol – Using predictive analytics to help prevent road accidents and save lives
6 percent reduction in traffic accident casualties
8.9 percent reduction in alcohol-related accidents
34 percent increase in driving-under-the-influence (DUI) arrests, making roads safer for all Solution components Software • IBM® Cognos® Business Intelligence • IBM SPSS® Modeler
The transformation: Previously, the THP worked in a largely reactive way, responding to accidents but lacking an aggregated view of when, where and why they happened. Today, armed with an IBM analytics solution that can predict future incidents based on correlations with past events, the THP can focus on likely hotspots and take preventative measures to reduce accident rates.
“The first full year the IBM predictive analytics solution was used in Tennessee was the second lowest traffic fatality year since 1963.” —Colonel Tracy Trott, Tennessee Highway Patrol
51
IBM Analytics
Why IBM ?
IBM Analytics
IBM a Leader in the Forrester Wave: Big Data Predictive Analytics Solutions, Q2 2015 “IBM assembles an impressive set of capabilities, putting predictive at the center. No matter how an organization wants to get started with predictive analytics, IBM has an option for them. The solution offers one of the most comprehensive set of capabilities to build models, conduct analysis, and deploy predictive applications: both on-premises and in the cloud. With customers deriving insights from data sets with scores of thousands of features, IBM’s predictive analytics has the power to take on truly big data and emerge with critical insights.” Source: Forrester Research Inc. “The Forrester Wave: Big Data Predictive Analytics Solutions, Q2 2015”, Mike Gualtieri & Rowan Curran, April 1, 2015
Source: Forrester Research Inc. “The Forrester Wave: Big Data Predictive Analytics Solutions, Q2 2015”, Mike Gualtieri & Rowan Curran, April 1, 2015 53
The Forrester Wave is copyrighted by Forrester Research, Inc. Forrester and Forrester Wave are trademarks of Forrester Research, Inc. The Forrester Wave is a graphical representation of Forrester's call on a market and is plotted using a detailed spreadsheet with exposed scores, weightings, and comments. Forrester does not endorse any vendor, product, or service depicted in the Forrester Wave. Information is based on best available resources. Opinions reflect judgment at the time and are subject to change.
IBM Analytics
Why IBM? Simplicity without sacrifice •
An easy-to-use visual approach to predictive analytics, with the depth and breadth of data access, manipulation and algorithms that meet the needs of a data scientist
Open and integrated •
The ability to access nearly any data source (including Hadoop for big data) and any data format (structured and unstructured); integration with marketplace-leading BI capabilities; and the combination of predictive analytics and business rules within multiple operational environments
Flexible deployment •
54
Predictive intelligence at scale to support a line-ofbusiness function with insight and to drive decisions across the enterprise
IBM Analytics
Resources
Explore ibm.co/predictive
Try
Learn
ibm.com/tryspss
www.BigDataUniversity.com
Course ID: 604
Engage https://developer.ibm.com/ predictiveanalytics/
56
