The IBM SPSS Predictive Analytics portfolio provides users with advanced statistical analysis that is
easy to use, flexible and scalable. With its ability to derive insights from both structured and
unstructured data along with machine learning algorithms, it puts predictive analytics into the hands
of every person and process in the organization.
The two major products within the portfolio are IBM SPSS Statistics and IBM SPSS Modeler. So,
which one is best suited for your needs? Simply put: IBM SPSS Statistics supports a top-down,
hypothesis testing approach to your data, while IBM SPSS Modeler exposes patterns and models
hidden in data through a bottom-up, hypothesis generation approach. In the sections below, I have
provided a more detailed look into the two products, their descriptions and their uses.
IBM SPSS Statistics
The world’s leading statistical software used to solve such business and research problems
by means of ad-hoc analysis and hypothesis testing
Description:
SPSS Statistics enables you to quickly dig deeper into your data, making it a much more effective tool than spreadsheets, databases, or standard multi dimensional tools for analytics. It excels at making sense of complex patterns and associations— enabling you to draw conclusions and make predictions.
Main Use:
Hypotheses testing approach
Use Case:
- Have a need for descriptive and predictive analytics
- Data is already collected for non-analytical purposes
- Need to create regular analytical reports
- Need to test data for statistical significance
Additional features:
- Advanced algorithms, procedures, and extensions that cover both statistical
and predictive analytics - A robust suite of data preparation features
- Reporting and job scheduling capabilities
User feedback:
“Ease of use without writing syntax, but also straightforward syntax when
needed. Online help resources. Can do some advanced statistical procedures.”
“More transparency and confidence, as we can always view the dataset in
totality, after each step of data transformation.”
“The learning curve to using this product is not steep. The program is
appropriate for those who do not have a lot of background in programming,
yet have to perform basic statistical analysis.”
“Custom tables and macros allow us to create useful reports quickly for a
broad audience.”
IBM SPSS Modeler
A data science platform that brings predictive intelligence to decisions made by
individuals, groups, systems and the enterprise
Description:
IBM SPSS Modeler is a leading visual data science and machine learning (ML)
solution that enables users to consolidate all types of data sets from dispersed
data sources across the organization. It helps in data preparation and
discovery, predictive analytics, model management and deployment, and
machine learning to monetize data assets.
Main Use:
Hypotheses generating approach
Use Case:
- Need to develop models that generate outcomes for operational decisions
- Need to combine data from many sources or database tables
- Analyzing/ querying data mostly on ad hoc basis. Modeler is more commonly
used for ‘pattern detection’ type problems than traditional reporting - Data originally collected from customer databases and flat files that were
originally collected by marketing, billing or CRM applications with analysis in
mind
Additional features:
- More than 30 base machine learning algorithms
- Enhanced support for several multithreaded analytical algorithms, including
Random Trees, Tree-AS, Generalized Linear Engine, Linear-AS, Linear Support
Vector Machine and Two-Step-AS clustering - Prepared data can be exported from IBM SPSS Statistics via XML to IBM SPSS
Modeler
User feedback:
“Automated modeling for classification, clustering, linear modeling, and
forecasting is very useful.”
“A lot of jobs that are stuck in Excel due to the huge numbers of rows are
tackled pretty quickly.”
“It’s very easy to use. The drag and drop feature makes it very easy when you
are building and testing the streams.”
To conclude, the main difference is the manual, user driven, top-down approach to data analysis
supported by IBM SPSS Statistics versus the data-driven, self-organizing, bottom-up approach to
data analysis (that works on very large data sets) supported by IBM SPSS Modeler. It should be noted
that both approaches can drive Predictive Analytics. Want to find out more about these products?
Join our IBM Analytics 101 Webinar by signing up at https://dataskill.com/analytics101