Naive-Bayes-Classifier-for-spam-filtering

Python based implementation of Naive Bayes Binary Classification algorithm from scratch for email spam detection. Pre-processing of the dataset was done using tokenisation and punctuation removal. Laplace Smoothing has been applied to handle ZeroDivisionError while computing probabilities. The program also uses 7 fold cross validation for the optimisation of the classifier model and achieves a best case accuracy of 72% on the test dataset.