SAS® Visual Data Mining and Machine Learning
Huge performance gains. Innovative algorithms.
One in-memory environment.
Interactive and programming interfaces. Innovative algorithms. Fast, in-memory processing. SAS Visual Data Mining and Machine Learning shatters barriers related to data volume and variety, limited analytical depth and computational bottlenecks. That means greater productivity – and faster, deeper insight.
Solve complex analytical problems faster.
Get breakthrough performance from predictive modeling and machine-learning capabilities running on SAS® Viya™, the latest addition to the SAS Platform. In-memory data persistence eliminates the need to load data multiple times during iterative analysis. Multiple users can collaborate easily, exploring the same raw data and building models simultaneously. You'll measure processing time for analytical modeling in seconds or minutes – not hours – and find solutions to challenging problems faster than ever.
Access powerful SAS® Analytics from an open ecosystem.
Don't know SAS code? No problem. Your open source programming skills are all you need to experience the power of SAS advanced analytics. Whether you’re using Python, Java, R or Lua, you can call SAS Analytics using your preferred language in our world-class, governed environment. And using REST APIs, you can add the power of SAS Analytics to your custom applications.
Explore multiple approaches quickly to find the optimal solution.
The superior performance of the distributed analytical engine and feature-rich building blocks for your machine-learning pipeline enable you to explore and compare multiple scenarios quickly and easily. Automated model tuning lets you test different scenarios in an integrated environment to identify the best-performing model and find accurate answers.
Boost data scientist productivity.
Data scientists and other analytical professionals will reap huge productivity gains from the solution's single, integrated in-memory environment. They can access and prepare data. Perform exploratory analysis. Build and compare machine-learning models. Create score code for implementing predictive models. And they can do all this faster than ever before.
Perform machine-learning tasks in a robust visual or programming interface.
Web-based visual and programming interfaces provide an intuitive environment for easily configuring common machine-learning steps. You can code projects using SAS or other languages, such as Python, R, Java or Lua. The SAS code is generated behind the scenes for later batch runs, editing and automation. You can also create advanced machine learning algorithms using a visual drag-and-drop interface without ever having to code.
- A flexible, web-based programming environment. Interactive, web-based visual and programming interfaces let you perform the most common machine-learning steps – from data prep to model building, assessment and scoring.
- Highly scalable, in-memory analytical processing. Optimized for multipass analytical computations, the analytics processing engine provides a secure, multiuser environment for concurrent access to data in memory. Built-in fault tolerance and memory management ensure that processes always finish.
- Powerful data manipulation and management. The same distributed, in-memory environment is used for preparing data for analytics – accessing data, joining tables, subsetting and filtering data, and creating the final table for machine-learning projects.
- Data exploration, feature engineering and dimension reduction. Descriptive statistics and powerful graphical programming let you find and fix data issues, quickly identify potential predictors, reduce large data set dimensions and easily create new features from original data.
- Open ecosystem. Call SAS algorithms using a variety of programming languages – such as Python, R, Java or Lua – without having to code in SAS. Or add SAS Analytics to your custom applications using public REST APIs.
- Modern statistical, data mining and machine-learning techniques. The solution includes powerful unsupervised- and supervised-learning algorithms, such as clustering, principal component analysis, linear and nonlinear regression, logistic regression, decision trees, random forests, gradient boosting, neural networks, factorization machines (including tensor factorization) and support vector machines.
- Automated model tuning. Automated, intelligent hyperparameter tuning lets you discover optimal model configurations. You can use 18 different model assessment metrics as your tuning objective, and evaluate using either a validation partition or the built-in cross-validation mechanism.
- Integrated text analytics. Integrated text analytics capabilities include powerful text preprocessing, natural language processing, topic detection and more.
- Model assessment and scoring. Test different modeling approaches in a single run and compare results of multiple supervised-learning algorithms with standardized tests to quickly identify champion models.
Read the SAS Visual Data Mining and Machine Learning fact sheet for more details.
SAS is a Leader in The Forrester Wave™: Predictive Analytics and Machine Learning Solutions, Q1 2017.
Learn about the five big misconceptions of machine learning.