fvilmos / gaussian_naive_bayes_classifier

Simple implementation of a Gaussian Naive Classifier in python

Gaussian Naive Bayes Classifier

This is a simple implementation of a Gaussian Naive Classifier in python. Part of the probabilistic classifers [1], can acheive high accuracy on the given classification tasks.

Below the classifier is trained on the iris dataset[3] x, than random samples are classified with it o.

How to use it?

1. clone the project
2. import in your project the utils/gaussian_naive_classifier.py
3. load and format the dataset for traning to have a layout of Sampple Nr x Labels + Features
4. create an instance for "GaussianNaiveBayesClassifier" object gnb = GaussianNaiveBayesClassifier(labels=labels, label_data_indx=0, feature_data_indx=[1,2,3,4], return_class_label=True)
5. train it with your data gnb.fit_data(data)
6. use a data point to predict the class gnb.predict(sample)

See details of a working example in naive_bayes.ipynb, uses the iris [3] dataset for classifications.

Resources

1. Naive Bayes classifier
2. Gaussian function
3. iris dataset, source and credits

/Enjoy.