#baddatascience search results
Today, I saw this code on LinkedIn. This person is trying to create a small multiple chart, but is using for loops with Matplotlib subplots. This is a overly-complicated way to do it. There's a better way ... What's the better way? #Python #baddatascience
Plot twist: Both decisions lead to questionable results in production #DataScienceMemes #MLFails #BadDataScience #TeachableMoments #ModelValidation #DataScientistLife #AIHumor
#GlobalPeaceIndex ranks USA 114th most peaceful nation; better move that E54thSt office in NYC IEP! visionofhumanity.org/app/uploads/20… #BadDataScience
What am I sick of? So, so many people justifying their hiring of only white men, or of hosting events, making projects, building communities that are only welcoming to white men — and then saying, well 85%. Can’t help it. No one else is around. Pipeline problem. #baddatascience
Classic example of distorted chart with a clear bias to favorable vote. Green bar #1 of 40% has the actual length of 45%. Green bars 2) and 3) of 54% has the length of 60%. #BadDataScience guys ...
.@YouTube's recommendation system is getting dumber every day. Do you use the same data science team as @indeed? #baddatascience
Business requirements meets data science constraints #baddatascience
I'm shocked, shocked to find that all the focus on consumer fingerprint reading was on avoiding false negatives.
WHY IS ONE PARAMETER VISUALIZED IN TWO DIMENSIONS??????? #badDataScience #DataIsUgly #unnecessarilyConfusing
@InsightBrief @drvickip @2morrowknight I can tell you from reading the results, your questions skewed the answers. #BadDataScience
"Anecdotally, I've heard.." is a euphemism for "I recognize the availability bias is strong here but I'm still gonna make this point anyway." #baddatascience
In my head mason lives eternally being screamed at by senior generals about the meaning of macguffins, #BADDATASCIENCE
My phone keeps recommending me football matches even tho I've never shown any interest in any. Is it just coz I'm male? #baddatascience
When your valuation model includes asking price as a primary input. #BadDataScience #PredictiveAnalytics…
Just for the record @LinkedIn, I am not @sriramk. I am not @sriramkri either. #BadDataScience
I COMPLETELY agree! That is the fundemental problem a core belief in the sanctity and authority of the numbers without learning the methodology behind the numbers #Baddatascience
Insane that an algorithm that predicts risk and influences incarceration is NOT publicly available? #BadDataScience
Data-driven "risk assessments" in the criminal justice system are racially biased. Amazing @propublica project. propublica.org/article/machin…
Plot twist: Both decisions lead to questionable results in production #DataScienceMemes #MLFails #BadDataScience #TeachableMoments #ModelValidation #DataScientistLife #AIHumor
Today, I saw this code on LinkedIn. This person is trying to create a small multiple chart, but is using for loops with Matplotlib subplots. This is a overly-complicated way to do it. There's a better way ... What's the better way? #Python #baddatascience
Some economist also put out a study that said women's skirt lengths affected prices. #baddatascience
Not just bullying? What weather maps? Sunny day in London? Does it cover the commuters from St Albans or Nuneaton? Or is there any evidence that commuters would rather avoid a rainy day than a sunny one? How do you factor in annual leave in this model? #BadDataScience
Reading through COVID mispredictions like the housing market crash(It's soaring higher than ever #littlescary) or the Eviction wave #Baddatascience shows me that you gotta check how hot the source maker is. #Aspensabitspicy #Leaninistoospicy #Heritageisonfire
In my head mason lives eternally being screamed at by senior generals about the meaning of macguffins, #BADDATASCIENCE
I COMPLETELY agree! That is the fundemental problem a core belief in the sanctity and authority of the numbers without learning the methodology behind the numbers #Baddatascience
Classic example of distorted chart with a clear bias to favorable vote. Green bar #1 of 40% has the actual length of 45%. Green bars 2) and 3) of 54% has the length of 60%. #BadDataScience guys ...
Could it be the changed how much #baddatascience they were using before? Or it could be natural #herdimmunity is not being suppressed as much as it was last week. I wonder if flattening the curve will kill more than... youtu.be/_41qqhjY3GI
"Anecdotally, I've heard.." is a euphemism for "I recognize the availability bias is strong here but I'm still gonna make this point anyway." #baddatascience
LOL @HUTechTweets "80% percent accuracy and with no racial bias" as if racial bias is the only thing biasef in your test set #BadDataScience #nope. This is dangerous pseudoscience, please stop making the world an insane dystopia harrisburgu.edu/hu-facial-reco…
This algorithm looks at how much a person spends on healthcare instead of looking at the reason why the person is in the hospital? #baddatascience Blame it on the algorithm...but actually #SkimmLife theskimm.com/2019/10/24/ski… via @theSkimm
WHY IS ONE PARAMETER VISUALIZED IN TWO DIMENSIONS??????? #badDataScience #DataIsUgly #unnecessarilyConfusing
What am I sick of? So, so many people justifying their hiring of only white men, or of hosting events, making projects, building communities that are only welcoming to white men — and then saying, well 85%. Can’t help it. No one else is around. Pipeline problem. #baddatascience
Ah, @Dexcom: where their approach to the algorithm (more data = worse calibration; therefore withhold data) appears to mirror their approach to #techsupport & #custservice (more info = worse outcomes; therefore withhold info). #baddatascience #T1D #CGM #failedsensors #nosleep
The naive classifier for predicting whether a person who is also a top Trump campaign official gets indicted probably has better real-world performance than whatever model is being abused below to imagine caravan figures. #BadDataScience
Stop The Caravan! This is about being a nation of laws. You must enter our country LEGALLY. When will it be enough? When the caravan is 100K, 500K, 2MM in size?
The longer I look at it the less sense it makes #BadDataScience #CursedGraphs #AI #ML #DataScience (Thanks/Damn you @glenndavidson)
This Harvard Business Review chart seems to have been made by someone who didn't know what most of the terms meant: hbr.org/2018/10/which-…
Here's what I'm doing tonight: Trying to replicate this work. I doubt I will be able to. #BadDataScience medium.com/activewizards-…
Today, I saw this code on LinkedIn. This person is trying to create a small multiple chart, but is using for loops with Matplotlib subplots. This is a overly-complicated way to do it. There's a better way ... What's the better way? #Python #baddatascience
#GlobalPeaceIndex ranks USA 114th most peaceful nation; better move that E54thSt office in NYC IEP! visionofhumanity.org/app/uploads/20… #BadDataScience
Plot twist: Both decisions lead to questionable results in production #DataScienceMemes #MLFails #BadDataScience #TeachableMoments #ModelValidation #DataScientistLife #AIHumor
Classic example of distorted chart with a clear bias to favorable vote. Green bar #1 of 40% has the actual length of 45%. Green bars 2) and 3) of 54% has the length of 60%. #BadDataScience guys ...
Something went wrong.
Something went wrong.
United States Trends
- 1. Knicks 13.4K posts
- 2. Shamet 2,918 posts
- 3. #AEWDynamite 21.8K posts
- 4. #Survivor49 3,945 posts
- 5. #CMAawards 5,533 posts
- 6. Brandon Williams 1,047 posts
- 7. Derik Queen 4,362 posts
- 8. Vooch 1,160 posts
- 9. Vucevic 6,248 posts
- 10. Labaron Philon 1,117 posts
- 11. Blazers 4,211 posts
- 12. FEMA 68.1K posts
- 13. NO CAP 14.9K posts
- 14. #AEWCollision 8,768 posts
- 15. #SeeRed N/A
- 16. Coby White 1,126 posts
- 17. Nany 2,048 posts
- 18. Naji Marshall N/A
- 19. Chisa 31.9K posts
- 20. FREE HAT 1,657 posts