OpenGenus publishes explainers on algorithms, data structures, machine learning, and computer science theory, written by a group of contributors.
HN commenters cite it as a reference for specific topics like ReLU, hash table complexity, and non-Turing-complete languages, and one questions its O(n) claim.
If you don't like sigmoids you might like ReLU
https://iq.opengenus.org/relu-activation/
For many learning tasks ReLU performs better than sigmoid.
My favorite use of sigmoid functions is
https://scikit-learn.org/stable/modules/calibration.html
where they are very good at turning arbitrary scores (say from a full-text search engine) into probabilities. For IBM Watson they tried a lot of things and found logistic functions dominated. Turning scores into probabilities was how Watson could decide if and when hitting the button would help win the game.
The big trouble with probabilities is that, potentially, every event is contingent on every other event and the joint probability distribution of all possible inputs and outputs is a huge dimensional space. In principle you could learn any function by sampling the joint probability distribution exhaustively but practically you can't get that much data. The miracle of machine learning is that the methods we use can guess at the joint probability distribution of inputs and outputs with only a limited sample.
If there was one great unsolved problem of the "old AI" namely expert systems it was doing logical reasoning over probability functions. It's not good enough to estimate that A has a 80% probability of being true, in general you need to estimate what the probability of A is if B is true and C is false. If the problem cooperates you can use half-baked methods to reason about uncertainty (like the MYCIN medical…
NIST Dictionary of Algorithms and Data Structures[0]
Is there a specific subject area(s) of application? aka big data analytics[a] / data mining and web analytics[b] / streaming applications[c] / other
Random picks with mix of tutorials/examples/explainations/resouces from search engine term "probabilistic data structures": [1][2][3][4], probabilistic DS vs. postgresql[4]
---
[0] : https://xlinux.nist.gov/dads/
[1] : https://www.geeksforgeeks.org/introduction-to-the-probabilistic-data-structure/
[2] : https://iq.opengenus.org/applications-of-different-data-structures/
[3] : https://dzone.com/articles/introduction-probabilistic-0
[4] : https://blog.devgenius.io/probabilistic-data-structures-1-definition-and-bloom-filter-431e5cbfb934
[5] : https://www.postgresql.eu/events/fosdem2020/sessions/session/2934/slides/279/pg-probabilistic-structs.pdf
----
[a] big data : https://www.researchgate.net/publication/335418069_Probabilistic_data_structures_for_big_data_analytics_A_comprehensive_review
[b] data mining / web analytics : https://highlyscalable.wordpress.com/2012/05/01/probabilistic-structures-web-analytics-data-mining/
[c] streaming : https://www.usaclouds.org/blog/streaming-with-probabilistic-data-structures-why-how
Not sure why the link is to the discourse forum... the article is easier to read at this link [0].
I don't know much about "GPU design concepts", but this doesn't seem like it explains an awful lot. It appears to be a collection of largely unexplained figures about rasterization and some vague unexplained diagrams about GPU pipelines.
[0] https://iq.opengenus.org/basic-graphics-processing-unit-gpu-design-concepts/
> Presumably, without an allow the API doesn't function at all
An incorrect presumption. The API works perfectly fine without the need for permissions.
This is not a new technique. There's a decent write up here [0].
Permissions are only required to play to the actual device, or record from the actual device. Recording from nothing and not playing is allowed.
[0] https://iq.opengenus.org/audio-fingerprinting/
Good question, 'power' I reserve for turing completeness (some aren't, such as SQL before recursive CTEs were added, and... well, here's a list https://iq.opengenus.org/non-turing-complete-programming-languages/)
For "how much crap do I have to write to get it to do something", I call that expressivity. Scala is more expressive than fractran (I guess...).
Surprised promotional prizes & swag were not mentioned.
- (1) For example, YC startup had API integration contest with a real battle axe as the prize:
https://www.wufoo.com/blog/win-a-real-battle-axe-in-the-wufoo-api-contest/
- (2) Then there’s the free developer swag promotions:
https://iq.opengenus.org/ways-to-earn-free-cool-developer-swag/amp/