Create accountLogin

Recent searches

No recent searches

Search options

Only available when logged in.
dmv.community is one of the many independent Mastodon servers you can use to participate in the fediverse.
A small regional Mastodon instance for those in the DC, Maryland, and Virginia areas. Local news, commentary, and conversation.

Administered by:

Server stats:

154
active users

dmv.community: AboutStatusProfiles directoryPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.7

#datascience

38 posts36 participants7 posts today
Public *
Henrik Schönemann

Friends don't let friends use iris, those flowers are not innocuous:
"Many people using iris will be unaware that it was first published in work by R A Fisher, a eugenicist with vile and harmful views on race. In fact, the iris dataset was originally published in the Annals of Eugenics. It is clear to me that knowingly using work that was itself used in pursuit of racist ideals is totally unacceptable."
meganstodel.com/posts/no-to-ir
#datascience #data

Megan Stodel · Stop using irisThe iris dataset is very widely used in the data science community, whether as a training aid, a tool for trying out new skills, or just a well-known set of numbers that can be used as background while demonstrating something in a blog.
Public
Eric R. Scott

I got some unfortunate news earlier this month: #UniversityOfArizona has decided to defund the group I work for, @cct-datascience.bsky.social, amidst their continuing budget crisis.

datascience.cct.arizona.edu/ne

It super sucks and means I'll likely have to drop down to 50% in May and I'll be looking for new, better supported, #rseng or #datascience positions, ideally still working with biologists in #rstats in some capacity. Remote or in #Tucson.

Let me know if you know of anything!

Data Science Team · Data Science Team seeking new projects after losing funding
Public
datatofu

I have previously mentioned software standards in passing.

The top-level standard is ISO/IEC Std 12207, Information Technology—Software Life Cycle Processes, which is the international standard that defines a life-cycle framework for developing and managing (ALL) software projects.

This standard was adopted in the United States as IEEE/EIA Std 12207, Information Technology—Software Life Cycle Processes.

Obviously, there are more

Tags: #ai #python #datascience #tech #linux #opensource

Public
Jürgen

No robust solution in sight. #LLM progress is stagnant?

“According to #OpenAI’s internal benchmarks, their newer models– o3 and o4 mini– hallucinate more often than older reasoning models like o1, o1-mini, and o3-mini, as well as traditional models such as GPT-4”

#AI #tech #technology #datascience

theleftshift.com/openai-admits

The Left Shift · OpenAI Admits Newer Models Hallucinate Even MoreIn a technical report, the company said “more research is needed” to explain why hallucinations increase as reasoning capabilities scale
Public
Miguel Afonso Caetano

"This paper advances the critical analysis of machine learning by placing it in direct relation with actuarial science as a way to further draw out their shared epistemic politics. The social studies of machine learning—along with work focused on other broad forms of algorithmic assessment, prediction, and scoring—tends to emphasize features of these systems that are decidedly actuarial in nature, and even deeply actuarial in origin. Yet, those technologies are almost never framed as actuarial and then fleshed out in that context or with that connection. Through discussions of the production of ground truth and politics of risk governance, I zero in on the bedrock relations of power-value-knowledge that are fundamental to, and constructed by, these technosciences and their regimes of authority and veracity in society. Analyzing both machine learning and actuarial science in the same frame gives us a unique vantage for understanding and grounding these technologies of governance. I conclude this theoretical analysis by arguing that contrary to their careful public performances of mechanical objectivity these technosciences are postmodern in their practices and politics."

journals.sagepub.com/doi/10.11

#DataScience#STS#Insurance
ExploreLive feeds
About