subject
Intro

Practical Deep Learning For Coders, Part 1

 with  Jeremy Howard

This 7-week course is designed for anyone with at least a year of coding experience, and some memory of high-school math. You will start with step one—learning how to get a GPU server online suitable for deep learning—and go all the way through to creating state of the art, highly practical, models for computer vision, natural language processing, and recommendation systems. There are around 20 hours of lessons, and you should plan to spend around 10 hours a week for 7 weeks to complete the material. The course is based on lessons recorded during the first certificate course at The Data Institute at USF. Part 2 will be taught at the Data Institute from Feb 27, 2017, and will be available online around May 2017.

Syllabus

  1. Recognizing cats and dogs
  2. Convolutional neural networks
  3. Why deep learning. Intro to convolutions
  4. CNN architecture basics. Avoiding over and under-fitting
  5. CNN/SGD in Excel. Pseudo-labeling. Collaborative filtering
  6. Intro to NLP, keras functional API, and RNNs
  7. Embeddings in Excel. Building RNNs
  8. Exotic CNN architecures. RNN from scratch
5 Student
reviews
Cost Free Online Course
Pace Self Paced
Subject Deep Learning
Institution fast.ai
Provider Independent
Language English
Hours 10 hours a week
Calendar 7 weeks long
+ Add to My Courses
Learn Digital Marketing udacity.com

Gain real-world experience running live campaigns as you learn from top experts.

Advertisement
Become a Data Scientist datacamp.com

Learn Python & R at your own pace. Start now for free!

Advertisement
FAQ View All
What are MOOCs?
MOOCs stand for Massive Open Online Courses. These are free online courses from universities around the world (eg. Stanford Harvard MIT) offered to anyone with an internet connection.
How do I register?
To register for a course, click on "Go to Class" button on the course page. This will take you to the providers website where you can register for the course.
How do these MOOCs or free online courses work?
MOOCs are designed for an online audience, teaching primarily through short (5-20 min.) pre recorded video lectures, that you watch on weekly schedule when convenient for you.  They also have student discussion forums, homework/assignments, and online quizzes or exams.

5 reviews for Practical Deep Learning For Coders, Part 1

Write a review
6 months ago
Eric Perbos is taking this course right now, spending 15 hours a week on it and found the course difficulty to be medium.
I've seen many different Machine Learning courses from "big-league" Udacity/Coursera & co to small YouTube players like SentDex or Siraj_Raval. Also completed the Udacity Data Analyst nanodegree with a Machine Learning module taught by a Google Brain scientist. This Fast.ai is a superb one, full with hands-on cases an Read More
I've seen many different Machine Learning courses from "big-league" Udacity/Coursera & co to small YouTube players like SentDex or Siraj_Raval. Also completed the Udacity Data Analyst nanodegree with a Machine Learning module taught by a Google Brain scientist.

This Fast.ai is a superb one, full with hands-on cases and support explanations of Neural Networks on Excel (Yes, Excel FFS !) by Jeremy Howard (ex-Kaggle CEO).

It relies on Python 2.7 + Theano + Keras + Jupyter Notebook. And Part II with Python 3.5 +Tensorflow + Keras is about to be released.

Each weekly lesson comes with:

- 90 mins YouTube class video, starting with review of last week assignment.

- Video timeline so you can access a precise part later -thank the intern-

- Fully transcript-ed class notes -thank again the intern-

- Jupyter Notebooks used in class

- List of required and optional online readings

- Assignments for next week.

Plus a full Wiki, see for yourself: http://wiki.fast.ai/index.php/Main_Page

Now there's a pitfall for beginners indeed: initially this course designed in Spring 2016 required the use of Amazon Web Services (AWS) for computing and that part was super-messy -and potentially expensive if you forgot to stop your instance-.

Now fast-forward to Spring 2017 where you can get an Nvidia GPU GTX 1050 with 4gb CUDA-ready around 150$ and skip the AWS totally to build your own Deep Learning server at home. Jeremy's intern even provides a script to install the full Fiesta (Python 2.7 + CUDA + Theano + Keras etc.) on Ubuntu 16.04. for you lazy buggers !

Note: if installing a dual-boot Windows/Ubuntu on your home rig is too much code for you, you're in the wrong field...
Was this review helpful to you? YES | NO
9 months ago
profile picture
Anonymous is taking this course right now.
This is really a hidden gem in a field that rapidly growing. Jeremy Howard does an excellent job of both walking through the basics (going as far as to prototype Stochastic Gradient Descent algorithms in Excel), and presenting state of the art results. I was surprised time and again when not only was he presenting ma Read More
This is really a hidden gem in a field that rapidly growing.

Jeremy Howard does an excellent job of both walking through the basics (going as far as to prototype Stochastic Gradient Descent algorithms in Excel), and presenting state of the art results. I was surprised time and again when not only was he presenting material developed within the last year, but even within the week the course was running. The courses use of Keras as a high level API for Theano/Tensorflow is great for abstracting and conceptualizing the architecture of the deep learning models.

You not only get a walkthrough of how to use AWS for GPU leveraged learning, but you practice on real life data through Kaggle competitions. I would strongly recommend this course to anyone looking to go from zero real world experience to competing with experts in the field.
Was this review helpful to you? YES | NO
6 months ago
Satish Kottapalli completed this course.
I haven't found a better course for hands on deep learning. Jeremy covers all the practical aspects of DL, and covers the theoretical aspects without being intimidating. Highly recommended. Especially for people without heavy ML/DL background, but would like to incorporate the latest DL techniques in their domain
Was this review helpful to you? YES | NO
6 months ago
Bibhash Thakur is taking this course right now.
The only course that takes this code-centric approach to deep learning. Highly recommended for anyone who don't understand intuitions from mathematical formulae.
Was this review helpful to you? YES | NO
8 months ago
profile picture
Mark dropped this course, spending 10 hours a week on it and found the course difficulty to be easy.
Extremely difficult to get install up and working. There are many "known issues" with the install video tutorial, but the video hasn't been remade and the known issues aren't collected in one place, resulting in hours wasted, scouring the forums to get install working. Speaking of the forums, the organization of thre Read More
Extremely difficult to get install up and working. There are many "known issues" with the install video tutorial, but the video hasn't been remade and the known issues aren't collected in one place, resulting in hours wasted, scouring the forums to get install working. Speaking of the forums, the organization of threads is poor and no one seems to be monitoring the forums for new questions -- most of my questions went unanswered. I have had much better luck with other online classes where people are moving through as a cohort and so are reading and answering each other's questions on the same week's content. There may be great material in here, but the barrier to entry is too high. I was forced to give up after week one.
Was this review helpful to you? YES | NO

Class Central

Get personalized course recommendations, track subjects and courses with reminders, and more.

Sign up for free