k-NN-classifier

Building a k-NN classifier engine for learning purpose

Reference: This project is completely based on the guidance provided in the book "Python Object-Oriented Programming Fourth Edition by Steven F.Lott and Dusty Phillips". This project is created purely for learning purpose.

This project is ideal for learning purposes because the Iris Classification is a simple well prepared problem. This helps one to focus more on best practice in software development. This project includes the following:

1. Unit and regression tests
2. Design patterns like strategy design pattern for choosing different distance algorithms
3. Python's concurrent.futures for running computations in multiple cores to achieve massive speed boosts during the grid search for parameters
4. CICD using GitHub Actions

The above diagram shows the UML diagram of the project.

Tool Usage

Project Setup

Setup the required env variables

export CLASSIFIER_HOME_DIR=</../../k-NN-classifier>

Run the classifier

python knn_classifier.py --raw_inp_data_path <path_to_dataset>