These are chat archives for FreeCodeCamp/DataScience

7th
Feb 2018
Alice Jiang
@becausealice2
Feb 07 2018 00:32
Yeah :( but it's worth trying out at least.
evaristoc
@evaristoc
Feb 07 2018 11:00

PEOPLE

For the analysis of the chatroom data I am using pandas. However I am finding pandas objects are using the half of the RAM memory and the code is relatively slow.

I have seen an increasing use of h2o.io in Kaggle. I have been in meetups mentioning the software but never used it.

Anyone with experience in h2o.io? Are there people here interested in learning some?

evaristoc
@evaristoc
Feb 07 2018 11:18

PEOPLE

Pareto:
Did you know that...?: about 20% of the people who has been participating in the main fCC chatroom are responsible of more than the 80% of the messages. I suggest to expect this trend to be usual in different social media platforms.
Josh Goldberg
@GoldbergData
Feb 07 2018 13:24
This is common. Power law stuff... @evaristoc
Eric Leung
@erictleung
Feb 07 2018 16:24
@evaristoc ah the Pareto principle (80/20 rule) still holds true. That's really cool to have observed that in the chat rooms!
Eric Leung
@erictleung
Feb 07 2018 18:44
evaristoc
@evaristoc
Feb 07 2018 21:29
@erictleung @GoldbergData :)
Alice Jiang
@becausealice2
Feb 07 2018 22:42
Quick update on deep fakes that I mentioned a few days ago, pornographic content made by deepfakes or other deep learning apps and algorithms are being banned by several big platforms they were commonly found on, and all platforms are explaining the bans saying the content is non-consensual.
As excited as I am by the power and availability of this kind of technology, I'm super stoked that the ethics of it is being taken seriously.
Hopefully we'll get more like Nic Cage in every movie, now :)