ID3

An implementation of ID3 decision tree algorithm in C. The program is implemented and tested under Ubuntu 18.04.

Usage

Compile with Makefile

To compile the scripts, go into the src/ directory and run:

cd src/
make all

To remove the intermediate files created during compiling process, run:

make clean

Run Program

After compilation, an executable program namely classifier will be created. This program can take in 3 arguments: seed number, treefile folder path, dataset path.

By default, if running the program without any arguments, the program will use a randomly generated seed number, read in adult+stretch.csv dataset in data/ folder, and create a file with inferred decision tree in the current directory.

./classifier

To specify a folder to save the decision tree, run with -t flag. Example:

./classifier -t /home/Documents/tree

Note: Please do not add an extra / at the end of the path.

To input a desired seed number (usually used when trying to reproduce a previous result), run with -s flag. Example:

./classifier -s 3509273525

To use a dataset as training set, run with -d flag. Example:

./classifier -d ../data/breast-cancer.csv

Note: The dataset is required to have the first row specifying feature names and to use , as delimiter.

Here is an example when passing in all three arguments:

./classifier -s 3509273525 -d data/breast-cancer.csv -t /home/Documents/tree