F
FloydHub Blog
blog.floydhub.com · Company blog · on HN since 2017
FloydHub's company blog publishes machine learning tutorials on deep learning, genetic algorithms, Naive Bayes, reinforcement learning, and image colorization.
What HN saysHN commenters share its articles on genetic algorithms, Naive Bayes, photo colorization, and deep learning courses, with several authors posting their own pieces.
39HN people shared it
7linked it in comments
194stories submitted
5.6Kstory points
5 years agolast shared
2017people per year2026
Most-shared posts
- 1Turning web design mockups into code with Deep Learning4 people · 581 points · first shared 2018 · 116 comments on HN
- 2Inverting facial recognition models4 people · 18 points · first shared 2019 · 14 comments on HN
- 3Colorizing and Restoring Old Images with Deep Learning4 people · 7 points · first shared 2018 · discussion on HN
- 4Introduction to Genetic Algorithms3 people · 265 points · first shared 2019 · 65 comments on HN
- 5Naïve Bayes for Machine Learning3 people · 243 points · first shared 2019 · 24 comments on HN
- 6Exploring DNA with Deep Learning3 people · 130 points · first shared 2019 · 13 comments on HN
- 7Practical guide to hyperparameters search for deep learning models3 people · 70 points · first shared 2018 · 2 comments on HN
- 8Intro to RL: Q-Learning3 people · 25 points · first shared 2019 · discussion on HN
- 9Multiprocessing vs. Threading in Python3 people · 22 points · first shared 2019 · 1 comments on HN
- 10Best Machine Learning Books (Updated for 2020)3 people · 16 points · first shared 2020 · discussion on HN
- 11Turning Design Mockups into Code with Deep Learning3 people · 4 points · first shared 2018 · discussion on HN
- 12Distilling knowledge from neural networks to build smaller and faster models2 people · 230 points · first shared 2019 · 16 comments on HN
- 13Generative Adversarial Networks – The Story So Far2 people · 215 points · first shared 2019 · 25 comments on HN
- 14I analyzed the history of “Who is Hiring?” threads2 people · 203 points · first shared 2019 · 74 comments on HN
- 15Teaching My Robot with TensorFlow2 people · 177 points · first shared 2018 · 13 comments on HN
What HN people wrote when they linked it
ReDeiPirati · 2020-03-21
Since I'm mostly into ML/DL, I would like to share an article that I wrote one year ago, maybe some of you may appreciate it: https://blog.floydhub.com/best-deep-learning-courses-updated-for-2019/ (planning to refresh the content in a couple of months).
philipkiely · 2019-06-03
I just published a piece with Floydhub, it has been a great experience!
Check out https://blog.floydhub.com/introduction-to-genetic-algorithms/
memexy · 2020-06-08
Yes, the idea is that the model is local to the user and runs in their browser (pushing the computation to the edge). Here's a project demonstrating pose estimation from Twilio's engineering blog: https://www.twilio.com/blog/pose-detection-video-tensorflow-js.
Re: bandwidth. Do you mean the weights? If so then yes, you'd have to think about distilling large models to smaller ones. There are techniques for doing this. Here is an explanation of distillation from Floyd Hub's engineering blog: https://blog.floydhub.com/knowledge-distillation/.
dtjohnnyb · 2019-11-15
Similar one I enjoyed in the Naive Bayes article from yesterday https://blog.floydhub.com/naive-bayes-for-machine-learning/
Multiple Choice: If you choose an answer to this question at random, what is the chance you will be correct?
A) 25%
B) 50%
C) 60%
D) 25%
houqp · 2017-12-16
Shameless plug, we also have a benchmark post for CPU/K80/V100 focusing on image tasks: https://blog.floydhub.com/benchmarking-floydhub-instances. In this case, V100 can be a lot more cost effect compared to CPU and K80. Our tensorflow environments are built from source with optimizations targeting our CPU instances.
deepnotderp · 2019-08-15
This is not true at all, and no offense, is a misunderstanding of how NNs can be used in other applications.
For one, what you are modeling itself does not need to be differentiable, only the network itself needs to be.
Second, using neural networks in combination with other techniques for program synthesis is an active area of research currently, and although it is currently at around 50 lines or so, your fundamental assertion here is wrong.
Third, there are a number of ways that deep learning could be leveraged in app development. The easiest way would be to make heuristic-type decisions around UI/UX or to build parts of said GUIs using existing code blocks. This has already been used to some extent in website design (e.g. https://arxiv.org/abs/1705.07962 and https://blog.floydhub.com/turning-design-mockups-into-code-with-deep-l… So it's certainly possible that it could be used in conjunction with templates to build common app types.
Now, this startup is clearly not doing that, but, that doesn't mean that it's a) impossible to leverage AI for app development and b) all of deep learning is "just curve fitting"