Pure Language Processing with Classification and Vector Areas

About this Course

In Course 1 of the Pure Language Processing Specialization, you’ll:

a) Carry out sentiment evaluation of tweets utilizing logistic regression after which naïve Bayes, b) Use vector house fashions to find relationships between phrases and use PCA to scale back the dimensionality of the vector house and visualize these relationships, and c) Write a easy English to French translation algorithm utilizing pre-computed phrase embeddings and locality-sensitive hashing to narrate phrases by way of approximate k-nearest neighbor search. By the tip of this Specialization, you’ll have designed NLP functions that carry out question-answering and sentiment evaluation, created instruments to translate languages and summarize textual content, and even constructed a chatbot! This Specialization is designed and taught by two specialists in NLP, machine studying, and deep studying. Younes Bensouda Mourri is an Teacher of AI at Stanford College who additionally helped construct the Deep Studying Specialization. Łukasz Kaiser is a Workers Analysis Scientist at Google Mind and the co-author of Tensorflow, the Tensor2Tensor and Trax libraries, and the Transformer paper.

WHAT YOU WILL LEARN

• Use logistic regression, naïve Bayes, and phrase vectors to implement sentiment evaluation, full analogies & translate phrases.

   

   

SKILLS YOU WILL GAIN

• Machine Translation
• Phrase Embeddings
• Locality-Delicate Hashing
• Sentiment Evaluation
• Vector Area Fashions

Syllabus – What you’ll be taught from this course

Content material Score96%(19,608 rankings)

WEEK

1

11 hours to finish

Sentiment Evaluation with Logistic Regression

Be taught to extract options from textual content into numerical vectors, then construct a binary classifier for tweets utilizing a logistic regression!

WEEK

2

7 hours to finish

Sentiment Evaluation with Naïve Bayes

Be taught the idea behind Bayes’ rule for conditional chances, then apply it towards constructing a Naive Bayes tweet classifier of your individual!

WEEK

3

8 hours to finish

Vector Area Fashions

Vector house fashions seize semantic which means and relationships between phrases. You’ll discover ways to create phrase vectors that seize dependencies between phrases, then visualize their relationships in two dimensions utilizing PCA.

WEEK

4

8 hours to finish

Machine Translation and Doc Search

Be taught to remodel phrase vectors and assign them to subsets utilizing locality delicate hashing, with the intention to carry out machine translation and doc search.