/ai_project_2018

This is a project designed for Artificial Intelligence (CS:4420) 2018, and the project is able to create a decision tree using F# and make decisions based on the tree.

Primary LanguageF#

AI PROJECT DECISION TREE

How to run our project

Install necessary libraries

Please use Visual Studio or whatever necessary to get F# libraries installed. For this project, we are using FSharp.Data and MathNet.Numerics. After that, change your reference paths on top of the Index.fsx file.

Change directory path

Please update your directory path on variable BASE_PATH so it can find its relative paths of the .names and .data files.

How to run

In Visual Studio Code, plase make sure the F# package is installed, then select all of Index.fsx code and press Alt + Enter

In Visual Studio, similar approach can be used. However, one may also press the run button to have a compiled executable running alone.

Generated files

Training decision tree

The generated JSON file called train_result.json contains the necessary information for predicting the 70% training data in hold-out tests.

Testing decision tree

The generated JSON file called test_result.json contains the information needed for predicting the 30% testing data in hold-out tests.