I am a researcher in deep learning, and have given courses on the theory. Recently I was approached to give the same course adding sections using some very concrete examples of deep learning for companies.

My audience will be industry managers never exposed to deep learning before. What are the most easily performed examples I could provide without having to develop millions of lines of code that can be of general interest to managers never exposed to deep learning before? These examples need to be quite concrete and easily applicable.

In particular, they look for examples that can have the label "big data" slapped on them.

I know about using ML for detecting digits, or all the other applications on kaggle, but they do not seem to care for those.


2 Answers 2


This seems like a matter of opinion to me, but here are some applications of machine learning that you seen very often in Silicon Valley:

  • Advertising: what to show whom and when? The data comes fast and furious so the engineering part is quite sophisticated.
  • Recommendation: similar to the advertising problem, except there are no advertisers, and the volume is slower, so you can afford to use more sophisticated machine algorithms.
  • Security: did somebody outside penetrate your system? did somebody inside abuse your system?
  • Text/audio/video processing: information retrieval on media. Requires finding good representations of the items using machine learning; e.g., fetch all pictures similar to the input, where the similarity can be construed in various ways, such as similarity of texture, style, color, content, etc.

These are not necessarily simple, but they're representative. If you want to give a demo, I suggest the last one. You could show them how to find similar images.

  • $\begingroup$ Processing any significant amount of video should get the "big data" label that the OP is looking for. E.g. cs.stanford.edu/people/karpathy/deepvideo $\endgroup$ Mar 5, 2016 at 19:09
  • $\begingroup$ How difficult is it to find examples in "traditional", solid industry(car making, construction, chemistry, etc.)? $\endgroup$ Mar 6, 2016 at 15:04
  • 1
    $\begingroup$ Not very. Anywhere there is prediction involved (yield, sales, etc.), there is a potential job for machine learning. If self-driving cars count you could come up with so many examples ... I will leave enumeration of examples from "traditional" industries to people more familiar with them. $\endgroup$
    – Emre
    Mar 6, 2016 at 19:26

The most important application is even cross-domain usable.

It's known as CLD.

You buy some expensive hardware, or rent some external expertise. And while you are waiting for the company administration to make a move and allow you to install and operate the hardware (which cannot not centrally administered, because the GPUs fall out of their usual active directory policy admin pattern) you have to make sure you have something fancy to show off. So you get a number of images from your corporate intranet, mostly with company logos. Lots of pictures. And then you feed all of them into a neural network using the inception pattern, and train this on your home gaming PC. Do not use the pretrained puppy-snail model! So after you have pretrained the network, you can now "make Google dream your company logo" in arbitrary images. so you generate a few of them to hang on your office walls while waiting for your systems to be approved. And when the C-level execs come by and want to see results while you still are not even started, you can give them some of this artwork to take home. They will all be impressed by your expertise of deep learning! And that is why I call this application the CLD, the c-level distractor.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.