Decision Tree

Overview

This project was developed during my Master's program in Software Architecture and Design, specifically as part of the Adaptive Software Systems course. It's a simple prototype for building and interacting with a Decision Tree classifier using the ID3 algorithm.

The goal was to gain practical experience implementing machine learning models and understanding decision-making logic through code. The result is a console-based tool with features for importing/exporting data, querying a decision tree, and visualizing the results—all built around a lightweight, testable architecture.

Now, a few years later, I’ve decided to revisit and continue working on this project—improving the decision tree implementation and extending it to support regression trees as well.

---

What's Inside

• A functional decision tree builder (ID3-based)
• Console interface with import/export/query support
• Example datasets
• Basic tree visualization
• Unit tests using xUnit
• No fancy UI—just clean, testable code

---

How It Works

The Algorithm

The decision tree is built using the ID3 algorithm, which selects features based on Information Gain. It recursively splits the dataset until either:

• All samples belong to the same class, or
• There's no more useful info to split on

The tree can be queried using feature values to return a classification result.

Building the Tree

BuildTree(data): Recursively splits data using feature with highest Information Gain.
Stop condition: Subset is homogeneous or no further gain.

Querying the Tree

Traversal starts at the root node using user-provided values until a leaf node (classification result) is reached.

---

File Format

To import data, use a semi-colon-separated .csv file with a Data tag separating headers from values. The last column should be the class label.

✅ Correct Format:

Swimming Suit;Water Temperature;Swim Preference
Data
None;Cold;No
Good;Warm;Yes

❌ Incorrect Format:

Swimming Suit,Water Temperature,Swim Preference
None,Cold,No

---

Console Commands

When you run the app, you'll see a terminal-based menu with the following options:

1. Import Data – Load your .csv file
2. Export Data – Manually input and save data
3. Query Tree – Enter feature values to get a prediction
4. Help – Shows usage tips and format guide
5. Display Menu – Redraw the main menu
6. Print Tree – Visualize the decision tree
7. Quit – Close the application

Tree Example (in console)

+- SWIMMING SUIT
   +- none -> no
   +- good -> WATER TEMPERATURE
      +- cold -> no
      +- warm -> yes

---

Example Datasets

Included (or tested with) several datasets, including:

• chess.csv
• swimpreference.csv
• social_network_ads.csv
• species.csv
• goingout.csv

Each one was adapted to match the expected file format and tested with the tree.

---

Running the Project

Requirements

• .NET 8

How to Run

• With Visual Studio 2022: Open and run
• Via CLI: Run dotnet run in the project directory

Unit Tests

• Built using xUnit
• Run with dotnet test or through Visual Studio’s Test Explorer

---

Notes

This is a prototype, so it's not production-ready—just a hands-on implementation for learning and exploration. It's meant to demonstrate the fundamentals of decision tree construction and traversal in a clear, testable way.

---

Project Checklist

Feature	Status
Decision Tree (ID3)	✅ Implemented
Regression Tree	❌ Not implemented
Console Interface	✅ Done
Tree Visualization	✅ Basic console view
Import/Export	✅ Supported
Clean Documentation	✅ Included

---

Open TODOs

• Migrate from .NET Core 3.1 to .NET 8
• Add workflow for this project
• Refactor the Decision Tree implementation for better readability and modularity
• Implement a Regression Tree functionality