About the assignemnt: The assignment focuses on unsupervised methods (K means) using PCA for Dimensionality Reduction and visualising the data. The goal is to apply k-means clustering in order to create clusters from the dataset - for fake and real news for given articles.

About the data - Fake.csv and True.csv - Contains list of articles considered as "fake" or "real" news. each record holds information on particular article.

Main parts:

1. Read CSV files and create one single dataframe **
2. Cluster the samples into two clusters - real and fake
3. Apply PCA and visualize the clustering in 2D or 3D
4. Performance Evaluation - Accuracy 5+6. Find optimal k - using Siloette and SSD scores according to different k values**
5. Apply PCA and K-means, Generate the features using tf-idf vectorizer