FICO® Decision Optimizer enables business analysts to discover optimized decision strategies that balance trade-offs between cost, risk and reward, while also factoring in economic and market conditions.
- It is a FICO Xpress Solution built using FICO Xpress Insight. It is an example of a Framework Solution that is fully configurable through its user interface - the underlying optimization model is fixed.
- Its advanced decision impact modeling, simulation and optimization techniques allows you to discover better decision strategies that balance trade-offs between cost, risk and reward, while also factoring in economic and market conditions.
- FICO Decision Optimizer v7.4 is now generally available for customer managed Cloud and On-Premises deployments. It will become generally available on FICO Analytic Cloud in the next month or so.
The Decision Optimizer v7.4 release brings new functionality including:
- Integrated Decision Tree Development Capabilities
- Users can now develop data-driven decision trees within Decision Optimizer (DO). There is no longer a need to use separate decision tree development software alongside DO.
- The primary use case is for the development of decision trees based on optimization scenario results. The treatment allocated by the optimization scenario is set as the tree development target, and a CaRT (Classification and Regression Tree) algorithm is used to build out a tree segmentation that best fits the target.This process allows business analysts to refine the optimization output into a consistent & explainable decision tree strategy that can easily be deployed in FICO applications.
- There is also an option to color code the different targets, to quickly visualize the purity of each end node in relation to the target, and help the user focus on nodes needing more analysis.
- It can also be used to support a range of requirements within DO such as: developing challenger decision tree strategies for use as inputs into Simulation Scenarios, creating new tree Decision Impact Model components, supporting segmentation analysis, assessing the criteria used to define current decisions, and editing or refining trees created by tree aware optimization.
- Stress Testing Scenarios
- Stress Testing Scenarios allows users to understand the impact of changes in assumptions on calculations and output metrics. For example, what if the bad rate is 5% higher than expected? This helps identify more robust decision strategies that will perform well in a range of competitive, economic or market conditions.
- Both optimization and simulation scenarios now support Stress Testing scenario exploration.
- A range of possible stressor values can be applied, as actual, additive or multiplicative values, with a sub-scenario created for each. This allows users to easily understand how changes in assumptions impact results.
- This release also includes a range of other usability enhancements focused on Treatment management, Validation Scenarios, Segmentation and Reporting
Get your hands on the software today with a free trial.