Machine Learning

A machine learning model for identifying 20 diverse languages as well as reporting ‘Other’ for languages on which model was not trained. The approach uses character (within word boundaries) level TF-IDF featurizer followed by a Multinomial Naive Bayes classifier.

A vectorized iterative implementation of the paper “Retrofitting Word Vectors to Semantic Lexicons” in which the vector space representations are further refined using relational information from the semantic lexicons.

Parallel implementation of SVM with cascading using MPI and Sequential Mimimal Optimization (SMO) using Cuda.