Post on 17-Apr-2018
IBM SPSS StatisticsProduct capabilities overview 3.3.2016
Analytics
© 2016 IBM Corporation2
Session agenda
IBM SPSS Statistics positioning4
IBM SPSS Statistics overview1
IBM SPSS Statistics usage2
Example use cases3
Analytics
© 2016 IBM Corporation3
Session agenda
IBM SPSS Statistics positioning4
IBM SPSS Statistics overview1
IBM SPSS Statistics usage2
Example use cases3
Analytics
© 2016 IBM Corporation4
IBM SPSS Statistics overview
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, top-down
approach to data analysis
Data mining involves
– the interrogation of the data
– determined by the method and
goal, rather than by the user
It is a data-driven, self-organizing,
bottom-up approach to data analysis that
works on very large data sets
Statistics Approach Modeling Approach
Both approaches drive predictive analytics
IBM SPSS Statistics IBM SPSS Modeler
“Statistical Modeling: The Two Cultures,” Leo Breiman, Statistical Science, 2001, Vol.16 (3), pp.199-231.
Analytics
© 2016 IBM Corporation5
IBM SPSS Statistics overview
• Hypothesis testing, advanced
statistical analysis
• Top-down approach
• Spreadsheet-like look & feel
• General environment for predictive
analytics and statistical analysis
• Well-suited for ad-hoc analysis
– Core descriptive statistical capabilities
– Advanced statistical functions
– Many types of regression
– Charting and mapping capabilities
– Tabular analysis & output
Analytics
© 2016 IBM Corporation6
IBM SPSS Statistics overview
Quickly understand large and complex
datasets using advanced statistical
procedures ensuring high accuracy to
drive quality decision-making
Reveal deeper insights and provide
better confidence intervals via
visualizations
Solve complex business and
reseach questions via means of
statistical analysis and assumption
validation
Programmability for advanced users that
leverages common statistical programming
languages in the market (Python, R)
Analytics
© 2016 IBM Corporation7
IBM SPSS Statistics overview
IBM SPSS Statistics software is used by a variety of customers to solve industry specific business issues to drive quality
decision-making. Methods like forecasting, analyzing trends and assumption validation can provide a robust, user friendly
platform to understand your data and solve complex business and research problems.
IBM SPSS Statistics helps organizations like Lloyds TSB, Kent State University and the IRS have expanded their markets,
improved research outcomes while ensuring regulatory compliance and managed risk (evaluate programs, prevent crimes
and assess loan risks).
Organizations have a core need to understand data and the statistical analyses applied to that data, and to quickly and
accurately analyze and interpret data to drive decision-making. In a variety of industries and applications, there is a need to
confirm (or deny) the existence of trends in data. IBM SPSS Statistics is the world’s leading statistical software used to
solve such business and research problems by means of ad-hoc analysis, hypothesis testing and predictive analytics.
Analytics
© 2016 IBM Corporation8
Session agenda
IBM SPSS Statistics positioning4
IBM SPSS Statistics overview1
IBM SPSS Statistics usage2
Example use cases3
Analytics
© 2016 IBM Corporation9
IBM SPSS Statistics usage SPSS Statistics is typically used by a statistician, who knows the technique to use (regression, decision tree, correlations,
etc.), and will:
1) Build descriptions from the data, e.g.
• How many customers are male, or from a certain age group
• Average amount per transaction
• How many students passed last year
• How many goals scored per match, on average
• What information is in my data?
2) Build inferences from the data
• Based on 1500 voters polled, who will win election?
• Based on past student retention rates, how many will finish the year?
• Based on revenue per quarter over X years, how much we can expect to bring in this quarter?
• Base on average points per game, what will the team (or player) score in the next game?
• Are our assumptions correct?
• Based on a sample of a population, what does the rest of the population look like?
• Based on a sample in time, what is the likely outcome in the future?
• What things actually have an affect on the results?
Analytics
© 2016 IBM Corporation10
IBM SPSS Statistics bundles
Statistics Base
Advanced Statistics
Custom Tables
Regression
StandardStatistics Base
Advanced Statistics
Custom Tables
Regression
Data Preparation
Missing Values
Categories
Decision Trees
Forecasting
Statistics Base
Advanced Statistics
Custom Tables
Regression
Data Preparation
Missing Values
Categories
Decision Trees
Forecasting
Bootstrapping
Conjoint
Exact Tests
Neural Networks
Direct Marketing
Complex Samples
Viz Designer
Amos
Sample Power
Professional Premium
Analytics
© 2016 IBM Corporation11
Session agenda
IBM SPSS Statistics positioning4
IBM SPSS Statistics overview1
IBM SPSS Statistics usage2
Example use cases3
Analytics
© 2016 IBM Corporation12
CRI now makes smarter clinical decisions that help patients and improve performance
The need
To advance clinical practice in the mental health sector,
Centerstone Research Institute (CRI) wanted to use emergent
analytics technologies to bridge the gap between researchers
and healthcare providers.
The solution
With the implementation of IBM predictive analytics software,
CRI built a framework that is able to predict which services are
likely to work best for which individual.
Real business results
Provided a potential 42% improvement in patient outcomes
Enabled cost savings of 58% in the cost per unit of outcome
change
Bridged the gap between researchers and physicians,
transforming the way mental health services are provided
Solution components:
IBM SPSS Statistics
IBM SPSS Modeler
"With our expertise in Big Data management, and IBM’s leading-edge analytics technologies, we’re well positioned to help
shift the paradigm for mental health services."
—Tom Doub, CEO, Centerstone Research
Institute
http://www-03.ibm.com/software/businesscasestudies/us/en/corp?synkey=J053479Y94979F29
Centerstone Research Institute
(CRI) is a not-for-profit research
organization dedicated to improving
the quality and effectiveness of care
for individuals with mental health and
addiction disorders.
Headquarters:
• Bloomington, Indiana
Patients served per year:
• 70, 000
Research Partners:
Harvard University Medical Center
Vanderbilt University
Indiana University
Northwestern University
University of Illinois of Chicago
Meharry Medical College
Janssen Pharmaceuticals
Telesage, Inc.
Analytics
© 2016 IBM Corporation13
Fondazione IRCCS Istituto Nazionale dei Tumori (INT) personalized cancer care
The need
Fondazione INT, a leading cancer and research center in Milan, wanted to improve patient care by
tailoring treatment approaches to specific individuals. The institute needed the ability to analyze
past treatments and cases, and combine that information with the patient’s personal statistics and
disease profile, to create a fact-based treatment plan for each patient. In addition, being able to analyze
overall outcome data would help the institute provide more cost-effective, efficient care for its patients.
The solution
The IBM solution proactively shows the physician statistics on similar clinical cases, possible
alternative treatments and predicted outcomes for each. Knowing this, physicians can ensure that
patients receive only the procedures they need.
Real business results
Avoids unnecessary treatment (estimated to be up to 60% of all treatment)
Increases the chances of successful outcomes by creating personalized treatments
Improves hospital performance, both clinical and operational, by streamlining processes and
lowering costs
"By providing our physicians with vital input on what worked best for patients with similar clinical characteristics, we can help improve treatment effectiveness and the final patient outcome.”
—Dr. Marco A. Pierotti, Fondazione INT
Solution components:
IBM SPSS Statistics
IBM SPSS Modeler
IBM Cognos BI
Analytics
© 2016 IBM Corporation14
Steno Diabetes Center ensured confidence in the validity of research
The need
Steno Diabetes Center is a medical research institution in Denmark, specializing in diabetes.
To break new ground in its research into diabetes and its complications, Steno needed to be
able to analyze complex data sets produced by research projects and clinical trials.
The solution
IBM SPSS Statistics provides data-handling and analysis capabilities that help Steno identify
significant factors in the development of disease and evaluate the effectiveness of
treatments.
Real business results
Increased ease of use for medical staff to combine data sets to perform complex
analyses, without help from statisticians or IT specialists
Trusted technology builds confidence in Steno’s published research
"The ability to handle and combine data sets from different cohorts of patients in IBM SPSS is very valuable... It’s a solid, mature and reliable platform, so we know we can trust it to deliver the correct results – and it gives the wider academic community more confidence in the validity of our studies.”
—Professor Peter Rossing, Chief Physician and Head of Research, Steno Diabetes
Center
Solution components:
IBM SPSS Statistics
Analytics
© 2016 IBM Corporation15
HUS analyzes sensordata to improve neonatal intensive care
The needA range of medical devices monitor the vital organs of neonatal infants within Neonatal Intensive CareUnits. Monitors attached to these devices display constantly changing data, usually at second-by-second intervals that caregivers must translate into actionable information. In addition to the Big Dataproblem, the early diagnosis of life threatening conditions is difficult because the clinical signs areusually vague and subtle until the condition is well established.
The solution
Applying machine learning algorithms to high frequency physiological data from patient monitors in
real-time to detect life threatening infections
Real business results
Life threatening conditions are detected sooner
Early warning gives caregivers the ability to proactively deal with complications
http://yle.fi/uutiset/hengenvaarallinen_ver
enmyrkytys_uhkaa_pikkukeskosia__wats
on-tekoalysta_etsitaan_turvaa/8316195
Analytics
© 2016 IBM Corporation16
Session agenda
IBM SPSS Statistics positioning4
IBM SPSS Statistics overview1
IBM SPSS Statistics usage2
Example use cases3
Analytics
© 2016 IBM Corporation17
IBM SPSS Statistics positioning
Statistics
The study of the collection, testing, and interpretation of data. Analysis is generally
performed on a subset of data (sample) and the analyst/researcher performs a
particular technique/s to validate a hypothesis.
Data mining
The analysis and organization and manipulation of observational data to find
relationships and patterns. Analysis is generally performed on all data available with
the intention to predict future outcomes. Multiple methods are attempted/used to
gain the most accurate/profitable model, that is tightly linked to the ultimate use case
at hand.
IBM SPSS Statistics
IBM SPSS Modeler
Analytics
© 2016 IBM Corporation18
Use case specific tooling for analytics
Questions IBM SPSS Statistics IBM SPSS Modeler
What are you trying to achieve? Hypothesis testing Goal Focused, support ongoing
business process
Who will be responsible for doing
it?
Analytics background (stats,
maths etc)
Data scientist (domain and
machine learning expertise)
What type of data will you be
using for the analysis?
Structured data Structured and unstructured data,
from several sources
Where does it reside? Excel, text files, small database Data warehouse, databases,
some flat files
How big is the data? Small data (usually 100’s or
1000’s, <1Mill)
Larger data (usually > 1000’s, to
Many millions)
How often do you want/need to
perform the analysis?
Ad-hoc Repeatable (weekly, monthly,
daily, real time)
What happens to the
results/output/score?
Reporting on results/findings (xls,
ppt, pdf)
Deploy results into some other
systems
Analytics
© 2016 IBM Corporation19