Innovative, Industry-Leading Trees
Angoss’ commitment to innovation is clear in our patented and patent pending Decision Trees and Strategy Trees which receive high praise from clients and industry analysts for being best-in-class.
Angoss Patented Decision Trees
Angoss Software is highly recognized for its unique patented, market-leading Decision Trees – which are at the core of the Angoss Software Suite, and provide an intuitive interface for building and exploring segments, discovering relationships between variables, performing variable reduction and much more.
“Angoss offers the best tool for Decision Trees that we have seen.” – Forrester Research
With an easy to use point-and-click interface, Angoss Decision Trees can be grown manually by creating custom splits or automatically using segmentation algorithms, or using a combination of both methods. This visual, real-time exploration aided by many usability features such as pan and zoom, variable search, and copying and pasting splits saves time for significant productivity and efficiency gains.
“Angoss received high marks from customers in flexibility and ease of use receiving praise for its decision tree interface.” – Hurwitz & Associates
Decision Trees Feature Highlights
- Discrete (categorical) and continuous dependent variables
- 4 algorithms based on improved variants of CHAID and CART algorithms
- Advanced algorithm control and user preferences to fine-tune tree growth, for example binary and n-ary tree options, value grouping types and open intervals
- Simple copy trees and splits (subtrees) within the same tree or between different trees
- Node Report that display node rules and statistics for every decision (terminal) node
- An Interactive Tree Map that is synchronized with the main tree view
- Dynamic node-monitoring charts to view the distribution of any variable within any node, synchronized with the main tree view
- Easy copying of tree images and underlying data to Microsoft Office® applications
- Automatic generation of deployment code in several languages: structured English, generic code, SAS, SPSS, SQL, Java, PMML and XML
Angoss Unique Strategy Trees
Unique, patent pending Angoss Strategy Trees are the first of their kind for building and deploying predictive strategies and are at the core of the Angoss Software Suite. Strategy Trees combine the usability of Angoss industry-leading Decision Trees with an intuitive, visual and collaborative approach for strategy design, authoring and validation workflow.
“Clients rave about Angoss’… strategy maps, which are clear differentiators.” – Srividya Sridharan, Forrester
Strategy Trees combine customer segments, scores, business rules, and calculations – applying user-defined treatments and actions to support decisioning and the development of business strategies, actions and optimization. For richer segmentation, Strategy Trees allow for the use of multiple target variables providing more feedback and control with key performance metric calculations for each node or segment. Automatic code generation for deployment eliminates manual coding errors.
A strategy in this context is defined as a combination of a predictive model represented by a generalized decision tree, a set of Key Performance Indicators (KPIs) calculated at every node and treatments. Treatments represent user-defined actions to be applied to specific segments (such as tree nodes).
“Angoss has the ability to transform other types of predictive models into a tree format and assign actions to create a strategy.”
Using the Strategy Tree framework, strategies can be developed both iteratively and interactively based on a range of “what-if” scenarios and forecasts created by the strategy author. The intelligent algorithm of the Angoss Decision Tree is used to assist in segmentation to optimize the splits.
Strategy Trees Feature Highlights
- Key Performance Indicators: Represents the metrics for each node that affect your decisions about the treatments to apply. Calculations can be specified in SQL syntax manually or using wizards for most common expression types. Calculations are updated dynamically as you change the segmentation
- Interactive Assignment: Interactively assigns treatments to the nodes identified in the segmentation process
- Multiple target variables: Provide richer segmentation, more feedback and control. Different dependent variables can be set at different tree nodes for developing complex strategies that address different objectives at different levels
- Node Report: Represents the segmentation as a spreadsheet and sorts segments by KPIs, making it easy to manage complex strategies with many segments or performance metrics
- Treatment Report: Monitors the scope and impact of the assigned treatments
- Strategy Validation and Monitoring: Compares data sets in the context of the performance metrics and segments defined by the strategy tree in order to identify segments that positively or negatively affect performance and monitor the change of indicators over time
- Strategy Optimization: Uses mixed-integer linear programming engine to automatically assign optimal treatment values—for example, proposed credit limit increase—under a set of linear constraints
- Strategy Deployment: Directly generates treatment lists or automatic conversion of a strategy tree to SQL, SAS or PMML program code