Build and retrain hundreds of predictive models across multiple segments – quickly and easily. Then automatically pick the best model for each segment. When modelers and statisticians work more efficiently, they have more time to spend unearthing valuable insights buried in granular segments. Such insights can reveal new opportunities, expose hidden risks, and fuel smarter, well-timed decisions.
Boost model building productivity.
Reap huge productivity gains by automating time-consuming model development processes – including data prep, variable transformation, predictor variable and algorithm selection, etc. SAS Factory Miner's easy-to-use web-based interface lets you build multiple models for each segment, and automatically identify the most accurate one.
Automate model development.
Choose the best segmentation strategy to solve your business problems, and jump-start your predictive modeling with a set of prebuilt model building templates that you can customize to fit your needs. Automated reporting and documentation make it easy to share best practices on model design and results across your organization.
Explore new ideas faster.
Apply machine learning and predictive analytic techniques to large, complex data sets, and get the results fast. If a model fails, you can try again quickly using different inputs or ideas. As variables change or new variables are found, you can test them without having to rebuild the entire data mining flow or challenge an existing set of algorithms.
Put models into operation quickly.
Deploy champion models in different production environments with just the click of a button. SAS Factory Miner automatically generates complete scoring code – including all necessary data prep and transformation steps. And retraining models is easy because all assets related to model development and deployment are centrally managed.
Demo & Screenshots
- Data preparation. Includes analytical data preparation capabilities.
- Customizable model templates. Provides out-of-the box model building templates that can be customized and shared across projects and users.
- Self-service machine learning techniques. Includes:
- Linear regression.
- Logistic regression.
- Decision trees.
- Random forests.
- Generalized linear models.
- Gradient boosting.
- Neural networks.
- Bayesian networks.
- Support vector machines.
- Champion model identification. Uses a variety of interactive assessment techniques.
- Model exception identification. Enables you to fine-tune models.
- Model retraining. Lets you retrain models over time using new data and variables.
- Scalable processing. Runs analytical procedures in a single machine, via grid computing or in-memory processing.
- Flexible model deployment. Lets you deploy models in database or in Hadoop to score new data using SAS Scoring Accelerator.
- Add-on to SAS Enterprise Miner. SAS Factory Miner runs as an add-on to SAS Enterprise Miner.
With a shortage of skills – a top challenge for analytics deployments – higher productivity for model builders is in high demand. SAS Factory Miner supports an industrialized approach to machine learning in a visual environment. This level of automation will benefit both business analysts and data scientists, minimizing the time to value for predictive models.