ML algorithms implemented in python from scratch. No machine learning library used.
Perform classification using Decision Tree on a simplified version of the Titanic dataset. Used information-gain as the splitting criterion.
Perform classification using Naive Bayes with Laplacian Smoothing.
Perform classification by implementing adaptive boosting decision tree stubs (Titanic dataset). Used information-gain as the splitting criterion.
Perform unsupervised clustering of the Iris Dataset, with 3 clusters. Evaluated the model by calculating the Jaccard Similarity Index for each cluster formed.