There’s a strong history of useful signals from single lead ecgs.
Detection of ecg anomalies(especially episodic ones with intermittent recording) was the subject of the physionet cardiac computing challenge almost 10years ago[0].
It’s amazing how far machine learning has come. I know teach a version of this challenge as a one day in class activity in my department’s physiology class. They actually get to train multiple models on a gpu cluster (and compare that to trying to train models on their laptops).
One thing we reinforce in the lesson is human vs. computer “interpretation”. They/clinicians can look at ecgs and make some sense of them. An LTSM is worse than random chance/a medical student. However moving to the frequency domain makes the LTSM more accurate than cardiologists, but neither they nor clinicians can “see” afib ina spectrograph. It’s a great way to talk about algorithmic versus human reasoning and illustrate that to students.
That then gets reinforced with other case studies of the ying and yang of human and machine decision making throughout our curriculum- like alpha fold working great until you ask it about a structure in the absence of oxygen, because that’s not in its training data.
> There’s a strong history of useful signals from single lead ecgs.
But to be clear, a single lead ECG requires two electrodes at a minimum and commonly a third as ground. So a single lead ECG will have minimum two cables attached to electrodes on the patient. The placement depends on which lead (eg lead I, lead II, etc) but there's always two minimum.
Basically by enabling drm (widevine). For the browsers/configurations that people knowledgable about browsers use for anything but streaming they'll force audio only mode and pretend that it's an acceptable solution.
The funniest part to me is these comments exclude something even more basic: joining a teams meeting on your phone. Notice the platforms mentioned ITA?
I regularly, as do many of the people here, join meetings from my phone. I often do so so I can squeeze a run in. I especially do so in the types of all hands or large meetings where I’m in listen only mode and things are shared that would be hard to trace back to any individual in the room.
I’m not carrying a second phone to take a picture of a slide - but I regularly take screenshots in those meetings to remind myself of something or to show someone when bitching.
The relevant xkcd here is decryption by wrench (538) - the problem being solved is not battling 1337 hackers, it’s herding normally distributed loan officers at a mid regional bank.
Because that is one of several goals. I heard a really interesting comment recently that concisely put what I find most dishonest about all this.
The opposite of DEI isn’t meritocracy it’s nepotism.
That is why you feel this way, the goal is to inhibit the success of those not part of the in group. The words bandied about about reverse racism and the like are just right wing propaganda.
It’s also easily abused…the parent post is a pretty solid example of how that happens. More than any individual action by the administration, decades of reinforcement and reification of this thinking in a major segment of society is what is going to doom us.
People celebrating their own destruction by spouting the propaganda they’ve been fed is somehow both terrifying and uniquely interesting to me.
They technically call it 'fringe benefits'. My university has four categories of fringe benefits:
Full
Limited
Partial
Grad Health
The only things it specifies are that partial includes social security and full includes life insurance. But given that whatever I set for a post doc/research scientist/etc. salary is the amount they are paid, I assume that everything else including payroll taxes are encompassed in that 1/3 extra for fringe.
> The guidance said that in the event of a request that “violates or doesn’t follow proper procedures”, employees were to contact Dorothy Aronson, the NSF’s chief information officer. “Do not give any indication that the request will be denied,” the guidance statement noted. Two members of DOGE, Luke Farritor and Zachary Terrell, were quickly given complete access to NSF grant-management systems despite statements in the guidance to staffers that they should initially receive read-only access.
I feel like “don’t make them follow policy” and “we’re going to lie about the access they have” is pretty telling as to whether there’s more behind this
Sometimes getting caught isn’t a bad thing. If you are trying to seed division between to groups, acting in a way that divides them - e.g., getting caught helping one side - is more effective than what you gain by not getting caught.
I struggle to see what Russia would gain with nlrb data, but getting caught “helping doge” furthers distrust between the two sides of our country - which is something they gain from
While I'm just guessing I'd think it would be better to wait until Ukraine is done and trump is out of office. Creating mistrust in Doge only helps Democrats
No, the two sides live in different information spheres.
This story will percolate up to many democrats who will be furious that Russia is “helping” “doge”.
Separately, it won’t (or will be dismissed as “overreacting” or “lying”) by republicans. They will see the democrats as overreacting and having trump derangement syndrome.
Meanwhile, the next doge encounter with an agency now brings greater fear of illicit acts for internal IT people and more controls for doge to demand are turned off creating more conflict within government function.
The sides believe in the evil and stupidity of the other will be further ossified. Meanwhile, Russia is effectively able to do espionage in a way where getting caught doesn’t diminish the value of the espionage work they are engaged in.
This is a great take but please don’t even dignify “trump derangement syndrome” by using it in conversation like this. That’s exactly what the people who created the term wanted it to be used for, ironically sowing further division.
I'm going to take a contrarian perspective to the theme of comments here...
There are currently very good uses for this and likely going to be more. There are increasing numbers of large generative AI models used in technical design work (e.g., semiconductor rules based design/validation, EUV mask design, design optimization). Many/most don't need to run all the time. Some have licensing that is based on length of time running, credits, etc. Some are just huge and intensive, but not run very often in the design glow. Many are run on the cloud but industrial customers are remiss to run them on someone else's cloud
Being able to have my GPU cluster/data center be running a ton of different and smaller models during the day or early in the design, and then be turned over to a full CFD or validation run as your office staff goes home seems to be to be useful. Especially if you are in anyway getting billed by your vendor based on run time or similar. It can mean a more flexible hardware investment. The use casae here is going to be Formula 1 teams, silicon vendors, etc. - not pure tech companies.
Detection of ecg anomalies(especially episodic ones with intermittent recording) was the subject of the physionet cardiac computing challenge almost 10years ago[0].
It’s amazing how far machine learning has come. I know teach a version of this challenge as a one day in class activity in my department’s physiology class. They actually get to train multiple models on a gpu cluster (and compare that to trying to train models on their laptops).
One thing we reinforce in the lesson is human vs. computer “interpretation”. They/clinicians can look at ecgs and make some sense of them. An LTSM is worse than random chance/a medical student. However moving to the frequency domain makes the LTSM more accurate than cardiologists, but neither they nor clinicians can “see” afib ina spectrograph. It’s a great way to talk about algorithmic versus human reasoning and illustrate that to students.
That then gets reinforced with other case studies of the ying and yang of human and machine decision making throughout our curriculum- like alpha fold working great until you ask it about a structure in the absence of oxygen, because that’s not in its training data.
[0] https://physionet.org/content/challenge-2017/1.0.0/
reply