Hi, My name is Surya Teja.Recently completed BTech from Bharath University and placed at Chubb limited for data scientist role.I need to learn deep learning and NLP.Can anyone guide me how to start deep learning and NLP.And any recommended course for deep learning and NLP??
Answer (1)
Hello
Deep Learning is known as a small branch of the Artificial Intelligence. With the help of a Deep Learning training, we can instruct our computers to learn for themselves in a way that gives us actionable outcomes.
Following are the Best Deep Learning Courses:
Best Overall Course: Deep Learning Nanodegree (Udacity)
Deep Learning Specialization (Coursera)
Complete Guide to TensorFlow for Deep Learning with Python (Udemy)
Deep Learning A-Z: Hands-On Artificial Neural Networks (Udemy)
An Introduction to Practical Deep Learning (Coursera)
Deep Learning, by 3Blue1Brown (YouTube)
Deep Learning: Recurrent Neural Networks in Python (Udemy)
Advanced AI: Deep Reinforcement Learning in Python (Udemy)
Deep Learning with Keras (Pluralsight)
Introduction to Deep Learning (Coursera)
The best Institutes for Deep Learning Courses in India are listed below:
Analytixlabs
Upgrad
Coursera
Udemy
Jigsaw Academy
Manipal ProLearn
Udacity
TalentSprint
Great Learning
ZekeLabs
The best way to learn about it is to go do it. Rob's suggestions of an intro course or a tutorial is a great way to get started, but really what you should do is try to solve some problem with NLP.
Here's how you might do that. Go download NLTK (Natural Language Toolkit). Then build some software that does something cool.
Here are some ideas:
Run a bunch of different POS taggers on some text and characterize the differences. Which is better for what domain? Do they make similar mistakes?
Build a sentiment analysis system and run it on a bunch of reviews, or tweets about something. See what mistakes it's making and figure out how to make it better. What features should you use that can help it do better? Find out if people liked the Interview, or whatever the movie du jour is!
Write a system to detect spam reviews on review sites. You can find a huge dataset here: Detect Fake Reviews and Reviewers or a smaller one here: http://myleott.com/op_spam/.
Build a system for some other language. Text segmentation in Chinese is fun and hard, but really pick any problem and apply it to another language, and you'll learn a huge amount about your assumptions about what works and what doesn't work.
None of these things will work well on your first try -- a key lesson in NLP, actually -- so the main thing you should focus on, once you get a baseline system working, is to figure out how to make your system better. That's 90+% of the effort in NLP.
Best of luck
Deep Learning is known as a small branch of the Artificial Intelligence. With the help of a Deep Learning training, we can instruct our computers to learn for themselves in a way that gives us actionable outcomes.
Following are the Best Deep Learning Courses:
Best Overall Course: Deep Learning Nanodegree (Udacity)
Deep Learning Specialization (Coursera)
Complete Guide to TensorFlow for Deep Learning with Python (Udemy)
Deep Learning A-Z: Hands-On Artificial Neural Networks (Udemy)
An Introduction to Practical Deep Learning (Coursera)
Deep Learning, by 3Blue1Brown (YouTube)
Deep Learning: Recurrent Neural Networks in Python (Udemy)
Advanced AI: Deep Reinforcement Learning in Python (Udemy)
Deep Learning with Keras (Pluralsight)
Introduction to Deep Learning (Coursera)
The best Institutes for Deep Learning Courses in India are listed below:
Analytixlabs
Upgrad
Coursera
Udemy
Jigsaw Academy
Manipal ProLearn
Udacity
TalentSprint
Great Learning
ZekeLabs
The best way to learn about it is to go do it. Rob's suggestions of an intro course or a tutorial is a great way to get started, but really what you should do is try to solve some problem with NLP.
Here's how you might do that. Go download NLTK (Natural Language Toolkit). Then build some software that does something cool.
Here are some ideas:
Run a bunch of different POS taggers on some text and characterize the differences. Which is better for what domain? Do they make similar mistakes?
Build a sentiment analysis system and run it on a bunch of reviews, or tweets about something. See what mistakes it's making and figure out how to make it better. What features should you use that can help it do better? Find out if people liked the Interview, or whatever the movie du jour is!
Write a system to detect spam reviews on review sites. You can find a huge dataset here: Detect Fake Reviews and Reviewers or a smaller one here: http://myleott.com/op_spam/.
Build a system for some other language. Text segmentation in Chinese is fun and hard, but really pick any problem and apply it to another language, and you'll learn a huge amount about your assumptions about what works and what doesn't work.
None of these things will work well on your first try -- a key lesson in NLP, actually -- so the main thing you should focus on, once you get a baseline system working, is to figure out how to make your system better. That's 90+% of the effort in NLP.
Best of luck
Comments (0)