rigorous testing & algorithm validation
documented and verified for regulatory compliance,
commercially viable, reliable, & proven for enterprise situations requiring the most accurate and trustworthy analytics
At SAS, analytical and statistical software is designed and written by highly specialized statisticians to make sure that the proper numeric algorithms are selected and implemented. In fact, of the employees working on our statistical products in SAS Foundation, 85 percent of the development and testing staff have advanced degrees and 64 percent of the employees with advanced degrees have PhDs in fields such as statistics, mathematics, and operations research.
Of the analysts who are working in our Advanced Analytics Lab, 98 percent have advanced degrees in fields such as statistics, mathematics, and operations research. Of those who have advanced degrees, 34 percent have PhDs.
Of the analytics developers and testers who work in the JMP division, 78 percent have advanced degrees, and 55 percent of those have PhDs.
The Quality Imperative: SAS Institute’s Commitment to Quality
Stability and reliability of SAS as a company;
history of innovation and leveraging customer input for s/w development
Wide variety of visualization and graphical capabilities
Extensive training, documentation and technical support
Stability and reliability of SAS as a company;
history of innovation and leveraging customer input for s/w development
Wide variety of visualization and graphical capabilities
Extensive training, documentation and technical support
Extensive set of data management capabilities
==============================
Data aggregation
Data quality
Data manipulation and transformation
Rpart is another popular package for calculating decision trees in R.
The Open Source Integration node can be set to PMML output mode for the widely used R models
lm (Linear Models), multinom (Multinomial Log-Linear Models),
glm (Generalized Linear Models),
rpart (Recursive Partitioning and Regression Trees),
nnet (Simple Neural Networks) and
kmeans (k-Means Clustering).
In PMML output mode, the Open Source Integration node translates the R model that is specified by the R object &EMR_MODEL into SAS DATA step code using PMML (Predictive Modeling Markup Language). The node then scores all imported data partitions with the generated SAS score code. The node automatically runs standard SAS Enterprise Miner assessments for supervised predictive models.
Ensemble models and doing model deployment is possible where R models can produce PMML output.
Note HP Data Mining includes HP Statistics.
HP Forecasting includes HP ETS.
The new HPCANDISC procedure performs high-performance canonical discriminant analysis.
The new HPFMM procedure performs high-performance finite mixture model analysis.
The new HPPRINCOMP procedure performs high-performance principal component analysis.
The new HPBNET procedure learns a Bayesian network from an input data set to create a predictive model in supervised data mining.
The new HPCLUS procedure enables you to read and write data in distributed form and to perform clustering and scoring in parallel.
The new HPSVM procedure executes the support vector machine (SVM) algorithm in multiple threads.
The experimental high-performance HPCDM procedure estimates a compound distribution model, which is the distribution of an aggregate loss that you expect to see in a given period of time.
The new HPCOPULA procedure is a high-performance version of the COPULA procedure, which enables you to simulate realizations of multivariate distributions by using the copula approach.
The new HPPANEL procedure is a high-performance version of the PANEL procedure, which analyzes a class of linear econometric panel data models.
analyzes a class of linear econometric panel data models.
The HPPANEL procedure provides the following four models:
one-way fixed-effects model
two-way fixed-effects model
one-way random-effects model
two-way random-effects model
The dashed line shows some usage but not the primary usage. Solid line shows primary usage.
Recommendation engine – both explicit and implicit recommendations
Recommendation engine – both explicit and implicit recommendations
The dashed line shows some usage but not the primary usage. Solid line shows primary usage.
* Talk about the fact that why VA is not greyed out.
Learn SAS
• Download and use SAS software anytime, anywhere. Once you download it, you can use SAS on a standalone PC, Mac or even a Linux workstation. And students can get it directly through SAS – no need to go through a professor or be enrolled in a class. SAS gives you a consistent user experience across all applications, whether you’re working on a class project or doing self-study. And you always have the ability to import and export all common PC file formats.
• Tap into a wide array of supportive teaching and learning resources. In addition to the fully functional software, SAS makes it easy for professors, students and researchers to learn, teach and stay connected to SAS:
free introductory videos to learn the basics of SAS programming and statistical analysis and get a jump on using features quickly.
interactive community for SAS academic users – including chat, support and introductory videos.
SAS Global Academic Program offers teaching and curriculum development materials for professors at no cost.
SAS Studio has several features to help reduce your programming time, including autocomplete for hundreds of SAS statements and procedures as well as built-in syntax help.
You can open a table in the table viewer, select which columns to display, and filter and sort the data.
SAS Studio generates SAS code through guided interaction with the user interface
You can use tasks in SAS Studio to generate code in a point and click interface