Order by request
Ab einer bestimmten Bestellmenge stellen wir Ihnen ein individuelles Angebot zusammen.
Dieser Artikel wird Ihrem Warenkorb beigelegt, wird jedoch nicht in der Berechnung berücksichtigt
Address a multitude of business problems and analytics requirements with a range of advanced algorithms and capabilities. SPSS Modeler Premium capabilities such as text analytics, entity analytics and social network analysis, along with automated modeling and preparation techniques, help you unlock insights from almost any type of data. SPSS Modeler Premium includes all the capabilities of SPSS Modeler Personal and Professional but is able to ddress a wider multitude of business problems and increase model accuracy with unstructured data.
IBM SPSS Modeler
Predictive Analytics in in a single solution
Today predictive analytics is a central key component in any modern IT-department of a company. Using predictive analytics to analyse and predict complex economic relationships gives your company a huge advantage and provides management with optimal operational decisions. Companies will be able to plan their decisions and strategies more effectively, predict risks more precisely and so on. IBM SPSS Predictive Analytics product line features broad and deep descriptive and predictive analytics, data preparation and automation and provides analytics of structured and unstructured data from virtually any source. The IBM SPSS Modeler is an extensive predictive analytics plattform, that brings predictive intelligence to individuals, groups, systems and companies for their decision making. The Modeler offers numerous advanced algorithms and procedures, including text analysis, entity analytics, decision management and optimization. Thus you can consistently make the right decisions - from the desktop or within operational systems.
IBM SPSS Modeler offers a wide range of analysis functions among them machine learning, automated modelling, ensemble modelling, simulation, geo spatial analysis and Big-Data algorithms.
IBM SPSS Modeler is available in different editions with adjusted fetaures to fit your needs. The following editions are available:
Downloads for the software IBM SPSS Modeler
Data Sheet | Helpful information about the IBM SPSS Modeler its applications and functions: read Data Sheet
White Paper | Use time and location-based intelligence to reveal hidden insights about your business, customers or constituents: read White Paper
Big Data with IBM SPSS Modeler | This document presents how to use IBM SPSS Modeler as an efficient source for accurate predictions: read Paper
- Users guide for installing single user licenses: ModelerAuthorizedUserInstall.pdf
- Administrator guide for installing single user licenses: ModelerAuthorizedUserAdmin.pdf
- Users guide for installing concurrent user licenses: ModelerConcurrentInstall.pdf
- Administrator guide for installing concurrent user licenses: ModelerConcurrentAdmin.pdf
System Requirements for the Software IBM SPSS Modeler
Details of the installation can be taken from the appropriate installation instroduction (see the tab 'Downloads').
|Windows®||Mac® OS X|
|Additional Requirements||Min. display resolution: 1024x768|
|Operating System||Windows 7, 8, 8.1, Windows 10 (32-/64-Bit)||Mac OS X 10.10 (Yosemite), 10.11 (El Capitan) (only 64-Bit!)|
|Min. CPU||Intel® Pentium® or Pentium-class processor, or higher (for 32-Bit Windows)
x64 (AMD 64 and EM64T) processor family (for 64-Bit Windows)
|Min. RAM||4 GB RAM|
|Disk Space||20 GB free disk space|
|Windows® Server||Linux® Server|
|Additional Requirements||Min. display resolution: 1024x768||Linux (64 bit) kernel 2.6.28-238.e15 or higher
FORTRAN version libgfortran.so.3
C++ Version libstdc++.so.6.0.10
|Operating System||Windows Server 2008 or 2012 (only 64-Bit!),||Red Hat® Enterprise Linux 6, 7, 7.1 (only 64-Bit!)
SUSE Linux Enterprise Server 11 (only 64-Bit!)
Ubuntu 14.10 (only 64-Bit!)
|Min. CPU||UNIX Hardware: PowerPC processor, 233MHz or better and IBM System p for IBM AIX
UltraSPARC II or better for Solaris
x64 (AMD 64 and EM64T) processor family or IBM System z for Linux 64-Bit
|Min. RAM||4 GB RAM|
|Disk Space||20 GB free disk space|
New functions in IBM SPSS Modeler
- Modeler Personal and Modeler Professional now running under MAC OS. (Modeler Premium will be added later!)
Big Data Algorithms
Some new algorithms have been added in previous versions (17.1) to the Modeler, however these could only be carried out in combination with the Analytic Server. Now all of these algorithms can be used directly in the Modeler without Analytic Server. There is also an improved time series algorithm. All these algorithms support parallel processing for the model building (model building will be done much faster - Big Data algorithms).
The following algorithms from earlier versions are now available in Modeler (without Analytic Server):
- Statistical methods: Linear-AS and GLE
- Linear Support Vector Machines
- Decision trees: Random Trees and Tree-AS (i.e. CHAID)
- Clustering algorithms: Two-Step-AS
Your benefits from the new algorithms: Multi-Threading
- Faster modeling with Big Data through parallel processing and more efficient use of hardware ressources
- All new algorithms are multithreaded even in the local Modeler (without Modeler Server, resp. Analytic Server)
- In previous versions of the Modeler, multithreaded algorithms relied on one of the above mentioned server
Your benefits from the new algorithms: Regularization
- Prevents “Overfitting” (inaccurate predictions of new data) by adjusting extreme and complexe parameters
- Building models without regularization often leads to excellent results just for the data on which the model was constructed, but not for new data.
- Available in GLE and Linear Support Vector Machines
Your benefits from the new algorithms: Automated Data Preperation
- Tree-AS and Linear Support Vector Machines preparing data automaticaly in the background
- Automated Data Preperation drasticaly reduces the amount of work you have to do as well as the error rate of a manuall data preperation
- Just three short examples for this feature: categorial fields with more than 12 occurrences will be merged (Default: <=12Bins), transformation of a date, resp. time field into a continious variable (i.e. birthday to age), empty spaces in string fields are 'trimmed'
Overview: Random Trees
- Random Trees are 'Ensemble Models', based on a diversity of decition trees (C&RT)
- The main goal of Random Trees are exact predictions of depending variables. So the identification of unknown correlations and patterns are not the focus.
Overview: New Time Series Algorithm
- 'Split' Modeling - different time series predictions are calculated for defined groups (using the split variable). Multithreaded and runnable in Analytic Server.
- If f.e. the split variable is the gender or the different stores of a retailer, then you can create forecasts based on time series.
Open Source: Python for Spark, Predictive Extensions
The IBM SPSS Modeler is capable of solely running Python for Spark (no Analytic Server needed). On-demand integration of Predictive Extensions are much more simplified. This shows that SPSS continues the way to integrate open source technologies in their system (see also R).
- Spark is an open source technology and very fast in the context of Big Data Analytics. Thanks to the in-memory technology of Spark it is actualy much faster than similiar techniques.
- Spark MLlib algorithms are accessable through Python for Spark and even for non-Hadoop data sources. Collaborative Filtering and Page Rank extensions are already available
- The Custom Dialog Builder now supports Python for Spark even without Analytic Server. The progrmmed code will be integrated into userfriendly GUIs and therefore enables the access to Spark functions even for 'non-programmer'.
- Predictive Extensions can be loaded within the Modeler directly so that algorithms can be used without any detours.
- New in Modeler: direct access to forums and the SPSS community
- The SPSS community is THE central contact point of SPSS users
- All technical infos & support, even Predictive Extensions
- Support via chat or e-mail, without a purchase of a license/ ICN/ Recorded Entitlement
- New DB2 on Z/OS In-Database algorithms. Five In-Database algorithms running in DB2 on Z/OS or IDAA (IBM DB2 Analytics Accelerator): Decision Tree, Regression Tree, K-Means, Naïve Bayes, Two-step