top of page

Exploit Machine Learning using VisiRule FastChart

VisiRule FastChart enables authors to start a VisiRule by using historical data.

Historical data often contains hidden trends and patterns. VisiRule FastChart can exploit machine learning algorithms to generate an initial VisiRule chart. Once exposed as a VisiRule chart, the logic can be executed, reviewed and refined, and deployed as an expert system.

What are Automated Decision Trees?

Automated decision trees are software programs that use data to create logical decisions. These trees are designed to take a set of conditions and outcomes, as well as a set of available actions, and then provide a structured way of choosing among the available options. By using data to create branches in the tree, the decisions can be made based on the best available information.

Why Use an Automated Decision Tree?

Automated Decision Trees provide an accurate, efficient and reliable method for making decisions based on data. They help to reduce the risk of human error and can be used in a variety of organizational settings. And in VisiRule, you can manually tweak or simplify the decision tree based on your own knowledge.

How Do Automated Decision Trees Work? 

Automated decision trees rely on four key components— data, algorithms, criteria, and conditions— to process input data or user requests. Data is first collected by an automated decision tree system either through a direct feed from a database or through manual entry by the user. Then, algorithms are used to sort the data based on specified criteria to identify the best path for a given solution. It’s important to note that these criteria should be clear and measurable in order for the automated decision tree to produce accurate results. Finally, conditions identify which rules need to be met in order for a certain path to be chosen. This helps ensure that decisions are made quickly and reliably based on provided parameters.

Induce Your Rules

You can use existing machine learning tools and packages to generate a decision tree.

This tree is then exported as PMML.

Knime-Decision-Tree
Imported-PMML-Decision-Tree

Import Tree Data

VisiRule FastChart will import your decision tree and convert it into a form which VisiRule can use.

VisiRule FastChart will identify the rules contained in the PMML decision tree and transform them into an internal format.

Imported Tree

VisiRule will rebuild a visual tree based on the imported data.

imported-PMML-decision-tree-Prolog-view
Imported-decision-tree-VisiRule-auto-generated

VisiRule Chart

Using the tree view, VisiRule FastChart will then create a brand new VisiRule chart - just as if you had drawn it.

Complex Charts

You can then extend, adapt your initial chart as you wish, combine it with other questions and logic or attach explanations and additional computations.

FastChart-Air-Handler-Decision-Tree
NDA-ChatBot
fever-medical-expert-system-chatbot

ChatBot Delivery

Once you have your VisiRule chart, you can deliver as a web-based expert system, or as an interactive chatbot

bottom of page