This project aims to compare the performance obtained using a linear Support Vector Machine model whose data was first processed through a Shortest Path kernel with the same SVM, this time with data also processed by two alternative Manifold Learning techniques: Isomap and Spectral Embedding.
