Naive-Bayes-Classifier

This project implements a Naive Bayes Classifier for predicting recurrence events in breast cancer patients. The classifier is trained on historical data, calculates the conditional probabilities of various features, and uses these probabilities to predict the class of new instances. This project does not use a ML library, the only libraries used is java.util and java.io

Features

age: Age of the patient (e.g., "10-19", "20-29", etc.)
menopause: Menopause status ("lt40", "ge40", "premeno")
tumor-size: Size of the tumor (e.g., "0-4", "5-9", etc.)
inv-nodes: Number of involved nodes (e.g., "0-2", "3-5", etc.)
node-caps: Presence of node caps ("yes", "no")
deg-malig: Degree of malignancy ("1", "2", "3")
breast: Breast side ("left", "right")
breast-quad: Breast quadrant ("left_up", "left_low", "right_up", "right_low", "central")
irradiat: Received irradiation ("yes", "no")

Class Labels

no-recurrence-events: No recurrence of cancer events
recurrence-events: Recurrence of cancer events

Project Structure

NaiveBayesClassifier.java: Main Java class implementing the Naive Bayes Classifier
Instance: Inner class representing a single data instance
data: Directory containing training and testing datasets

Installation:

Clone the repository:

git clone https://github.com/HugoBlair/Naive-Bayes-Classifier.git

cd path/to/Naive-Bayes-Classifier

Compile the Java code:

javac NaiveBayesClassifier.java

Run the classifier:

java NaiveBayesClassifier breast_cancer_testing.csv breast_cancer_training.csv