Natural Language Processing with Classification and Vector Spaces
"Natural Language Processing with Classification and Vector Spaces"
Although I already knew much of the contents taught in this certification, I was able to learn some new practical implementation tips and some new interesting concepts; in particular, I
- #sentiment_analysis under the supervised learning paradigm, logistic regression, data pre-processing, and feature extraction,
- Error analysis, conditional probabilities. Bayes rule, Naive Bayes inference, log-likelihood optimization, and Laplace smoothing.
- Covariance and co-occurrence matrices, vector representations, similarity, vector spaces, and dimensionality reduction through #PCA.
- Hash functions and tables in NLP, approximate #KNN, locality sensitive hashing, document search, and one basic approach to #machine_translation.
Coming from a more mathematical background, I was slightly disappointed at the mathematical rigorousness, but from a practical industry point of view, I found this to be an excellent accelerated introduction. Definitely recommend!
#machinelearning #nlp #coursera #deeplearningai #ai #datascience