flo_hu

8 ms·

1.
Being a minority in fancy coding land: a Windows user(blog.esciencecenter.nl)
2 points by flo_hu 6y ago|1 comments
2.
When to use Machine Learning for Research – Flowchart Guide to Help You Decide(blog.esciencecenter.nl)
1 points by flo_hu 6y ago|0 comments
3.
Mcfly 3.0 – An easy-to-use tool for deep learning for time series classification(blog.esciencecenter.nl)
1 points by flo_hu 6y ago|0 comments
4.
Monopolizing AI – larger models increase divide between big Tech and the rest(blog.esciencecenter.nl)
1 points by flo_hu 7y ago|0 comments
5.
2020 – and AI has still not taken over(medium.com)
1 points by flo_hu 7y ago|0 comments
6.
Highlights from the European Conference on Machine Learning 2019(blog.esciencecenter.nl)
2 points by flo_hu 7y ago|0 comments
7.
by flo_hu 7y ago
I removed my blog post from medium's distribution. So it should now be freely accessible! https://blog.esciencecenter.nl/king-man-woman-king-9a7fd2935...
8.
by flo_hu 7y ago
There's several use-cases listed where Markdown could be used for, such as paper writing, presentation etc (I'm not sure I am going to use it for the later in the near future...). I find that one of the most interesting potential
9.
Markdown can do much more than GitHub readmes(blog.esciencecenter.nl)
8 points by flo_hu 7y ago|1 comments
10.
by flo_hu 7y ago
No. That's of course still "King" :) (but sure one could also pick queen, prince, royal form the list...) Just tested it here: http://vectors.nlpl.eu/explore/embeddings/en/calculator/# And
11.
by flo_hu 7y ago
That's pretty much correct. You would typically calculate a vector for "King-Man+Woman" and then do a query on this based on a cosine distance (or similar measure) over the entire vocabulary. The query would give you a ranked
12.
by flo_hu 7y ago
I would also agree with you that it is fine to add additional rules to improve the outcome, but than it shouldn't be made clear in the way the result is presented (as you say, that rarely happens in intro-level tutorials/courses).
13.
by flo_hu 7y ago
Ok, thanks a lot for bringing this up! I will have a closer look at that.
14.
by flo_hu 7y ago
Good point! I would see this rather as yet another argument for why you should simply give the actual output of the NLP algorithm. So if people actually do the calculation King-Man+Woman and it comes closest to King, than they should report
15.
by flo_hu 7y ago
Exactly! I think that was part of the problem for many of the examples that turn out to not-really-work. People pretend they let the work do by an algorithm, but then hand-pick from a list of somewhat close candidates. Which of course happe
16.
by flo_hu 7y ago
I'm fine with the free lunch thing. But here the cheating is done on the level of how people present the capabilities of the tool. If you ask the algorithm how "SHE is to LOVELY as HE is to X", the reported answer (Bolukbasi
17.
King – Man + Woman = King?(medium.com)
256 points by flo_hu 7y ago|103 comments
18.
Could it be that self-driving cars are an outdated dream?(medium.com)
4 points by flo_hu 7y ago|1 comments
19.
by flo_hu 7y ago
Despite all great advances in deep learning and big data, scientific research often is more about getting the most from very little data.
20.
Scientific research: big questions – small data(blog.esciencecenter.nl)
2 points by flo_hu 7y ago|1 comments