3 ms·

I did this course, but couldn't finish all the assignments. Loved it. Please note that this is a convolutional neural networks course, not computer vision as su

I did this course, but couldn't finish all the assignments. Loved it. Please note that this is a convolutional neural networks course, not computer vision as such. From what I know computer vision encompasses a variety of non machine learning based algorithms, which are not covered in this course.

speedplane 9y ago
As an engineer, it's difficult to know when to use deep learning and when to use more classical algorithms. Often, you have to try both and see which is better (twice the work, hooray!). The classical algorithms often are very understandable, and you can reason with what's going on and figure out what is breaking. Deep learning is so much harder, e.g., are my hyper parameters bad, or do I need another 30 GPUs running for a week?
- amelius 9y ago
  Imho, deep learning has little to do with engineering, and more with guessing, hoping and praying. But it seems you can often get something to work if you do those three things hard enough.
  - deepGem 9y ago
    This has been my experience as well :). Deep learning is a lot of random guesswork, trial and error. I am almost always in the 'brute force' mode. However, in this course, you learn more about the fundamentals of convolutions and backprop. You have to implement your own backprop - not sure of what use that is, given that it's a one line code in TensorFlow
    - amelius 9y ago
      Studying how to write your own back propagation algorithm could be useful if you're a deep learning researcher. But for most people it would be like studying semiconductor physics if you only want to write software.
    - speedplane 9y ago
      I watched a great video on Tensorflow (link below). It mostly introduces very basic deep learning concepts, but there are a few key moments in the 2 hour+ video, where he explains what to do if something goes wrong. It's definitely not a "science", but with enough experience in deep learning, you can intuit what's going on inside the black box, and there are best practices on what to try next. For example, he goes through a few examples where a neural net has too many weights, or too little data or improperly connected nodes. All three result in problems, but the problems exhibit themselves in slightly different ways and with expertise you can start identifying them. https://www.youtube.com/watch?v=vq2nnJ4g6N0 https://www.youtube.com/watch?v=vq2nnJ4g6N0

speedplane 9y ago

As an engineer, it's difficult to know when to use deep learning and when to use more classical algorithms. Often, you have to try both and see which is better (twice the work, hooray!). The classical algorithms often are very understandable, and you can reason with what's going on and figure out what is breaking. Deep learning is so much harder, e.g., are my hyper parameters bad, or do I need another 30 GPUs running for a week?

amelius 9y ago
Imho, deep learning has little to do with engineering, and more with guessing, hoping and praying. But it seems you can often get something to work if you do those three things hard enough.
- deepGem 9y ago
  This has been my experience as well :). Deep learning is a lot of random guesswork, trial and error. I am almost always in the 'brute force' mode. However, in this course, you learn more about the fundamentals of convolutions and backprop. You have to implement your own backprop - not sure of what use that is, given that it's a one line code in TensorFlow
  - amelius 9y ago
    Studying how to write your own back propagation algorithm could be useful if you're a deep learning researcher. But for most people it would be like studying semiconductor physics if you only want to write software.
  - speedplane 9y ago
    I watched a great video on Tensorflow (link below). It mostly introduces very basic deep learning concepts, but there are a few key moments in the 2 hour+ video, where he explains what to do if something goes wrong. It's definitely not a "science", but with enough experience in deep learning, you can intuit what's going on inside the black box, and there are best practices on what to try next. For example, he goes through a few examples where a neural net has too many weights, or too little data or improperly connected nodes. All three result in problems, but the problems exhibit themselves in slightly different ways and with expertise you can start identifying them. https://www.youtube.com/watch?v=vq2nnJ4g6N0 https://www.youtube.com/watch?v=vq2nnJ4g6N0