SAS^® ADAPTIVE LEARNING AND INTELLIGENT AGENT SYSTEM

Use machine learning to automatically detect rare events such as fraud with minimal interaction.

Gartner positions SAS as a Leader in the Magic Quadrant for Data Science and Machine-Learning Platforms 2019

Combining both supervised and unsupervised machine learning to understand past anomalies, adapt, and retrain to predict future events.

SAS Adaptive Learning and Intelligent Agent System showing model scorecard on desktop monitor

Anomaly detection

Finds fraud and other rare events using state-of-the-art supervised and unsupervised machine learning algorithms and proven best practices.

Transparency

Takes the mystery out of machine learning with a model and record scorecard that delivers narrative, text and visuals for describing and interpreting models.

Easy-to-use interface

Enables users untrained in data science to create models for detecting rare events.

Adaptive learning

Creates an adaptive learning system that continually improves its ability to detect fraud and react rapidly to the emergence of new fraud patterns via integration with SAS Visual Investigator.

Automated integration

Automates integration with SAS Visual Investigator, saving you the time and expense of developing systems for scoring, alerting, triaging, refreshing training data and retraining models.

Extend model creation beyond data scientists.

SAS Adaptive Learning and Intelligent Agent System showing model properties on desktop monitor

Enhance productivity throughout the analytics life cycle.

SAS Adaptive Learning and Intelligent Agent System automates a major portion of the work required for model training and retraining. An easy-to-use interface enables you to expand the task of creating and deploying models tuned to detect rare events to a broader community of users with limited data science skills. This allows data scientists to focus their time on other strategic areas.

Automatically generate scores and alerts.

The solution easily integrates with SAS Visual Investigator. Once a model is deployed, jobs are created automatically to generate scores and alerts, which are fed into SAS Visual Investigator. Triaged alerts are fed back into the data so a new model can be trained. The solution supports multiple versions of a model, enabling users to compare models and tune new models before deployment. Automation ensures that models are retrained and refreshed more frequently to detect the latest fraud patterns.

SAS Adaptive Learning and Intelligent Agent System showing sample output on desktop monitor

SAS Adaptive Learning and Intelligent Agent System showing results on desktop monitor

Reduce false positives through accurate detection.

SAS Adaptive Learning and Intelligent Agent System combines state-of-the-art supervised and unsupervised machine learning algorithms, best practices and other fine-tuned techniques to accurately detect rare events and reduce the number of false positives.

Train models using the latest AI and machine learning techniques.

Users have the option to train a model using unsupervised learning, supervised learning or both. With unsupervised learning, it is now possible to train a model using data with no target (unlabeled). This model detects and alerts on anomalies in the data. SAS Visual Investigator can be used to investigate and disposition the anomaly alerts. Disposition results are automatically fed back into the training data to provide a target (label) for supervised learning.