Visual Workflow Environment
Build workflows with drag-and-drop interactive blocks to provide the perfect combination of low-level data engineering facilities for retrieving, blending, and preparing data for analysis, along with machine learning features that let you build, explore, and validate reproducible predictive models. Enhance workflows with programmable blocks coded in the languages of SAS, SQL, Python, and R.
Robust Coding Environment
Use a modern integrated development environment (IDE) to create, maintain, and run programs, and to explore data, results, and logs. Analytics Workbench’s coding environment focuses on SAS language programming but also lets users incorporate SQL, Python, and R code within SAS language programs, which can easily exchange data between Python, R, SQL, and SAS language modules.
Simple Data Discovery Capabilities
Analytics Workbench offers a robust range of functions that empower users to fully understand their source data and uncover new insights, including profiling, automated quality checking, validation, and automatic variable reporting.
No-Code Machine Learning Model Development
Analytics Workbench features machine learning support for supervised and unsupervised learning, including decision trees, clustering, regression analysis and neural networks. Explore, build, and test machine learning models with workflow blocks and automatically generate error-free code for production use.
Quickly Compare Model Performance
Build and validate different types of models against the same test data and then use Analytics Workbench’s no-code model comparison tool to identify the best model for you with comparison charts that include Receiver Operating Characteristics (ROC), Kolmogorov–Smirnov (KS), cumulative gain, and lift.
Use our simple visual development tools to build predictive, behavioral, and application scorecards to help with variable selection, training, evaluation, and model validation. Automatically extract error-free and ready to deploy scorecard code for use in production.