#Statistical Modelling
Explore tagged Tumblr posts
Text
Maximizing the Benefits of Risk Scoring and Classification in Forensic Analytics
Unlock the full potential of your forensic analytics with risk scoring and classification approaches. Learn how they can enhance fraud detection, improve regulatory compliance, optimize ICT systems and operations, analyze transactions and crime.
Unlock the Power of Forensic Analytics with Risk Scoring and Classification Forensic analytics plays a crucial role in many areas of business and government operations. From detecting and preventing fraud, to ensuring regulatory compliance and improving operations, to analyzing transactional data and detecting crime, the use of risk scoring and classification approaches can greatly enhance the…
View On WordPress
1 note
·
View note
Photo
Source http://smbc-comics.com/comic/take-it-off
642 notes
·
View notes
Note
White Apple Family does also make the thing with equestria girls big mac into "how do you whitewash a white guy", so there's that
we had blacked out eqg big mac until receiving this ask. god. im not sure we ever totally internalized that that guy was big mac like, emotionally
#ask#anonymous#Same way they whitewashed literally everyone else i presume#I mean ymmv on their individual ethnicities but i’m pretty sure they did it— again—#to literally every teen with an existing horse character model. So… statistically?
9 notes
·
View notes
Text
"i like to packbond with everything" is um. a really fucking concerning statement to make when discussing AI, especially the question of if AI can reach sentience. that really should be the very first alarm bell you learn to watch for when you get into studying AI 😬
#i never formally went into AI research due to a mixture of disdain for the majority of commercial AI products + undergrad almost killing me#but ive kept up more or less with where things are at and like#a) we are still using a statistical model#b) we still dont understand how the brain utilizes its heuristics to switch between them#which is first of all the main failure of AI to generalize (or even just not hyper specialize)#when you compare it to the human brain like#the brain has all kinds of weird statistical models (approximately) right#these are heuristics#just ways of determining is thing x#we can more or less do this task with AI!#the PROBLEM is that the brain then can determine Which Heuristic Is Most Appropriate (more or less)#we have no fucking clue how it does that#or really how the feedback loop of learning actually works for meat#the statistical model was sort of our first/most popular attempt to try simulating a feedback learning loop in the brain#and like. mixed results! again! AI can't generalize and we dont even know how the brain works properly#so my personal opinion is that we cannot simulate the human brain any better with current approaches to the problem#and like. if we cant even simulate the bare bones functionality of the brain.#how the everliving fuck do you think it can simulate thought or existence?#we dont know how that works either but like its clearly generated by SOMETHING in the brains functionality#without that you got no hope of a sentient computer#flat out. it isnt happening.#what we DO know though is that the human brain is an anthropomorphizing machine#it can interpret anything as alive or humanlike#we know however this is a fucking trap in regards to intellectual pursuits#ergo: talking about pack bonding and AI study in the same breath?#you are victim to one of the human brain's fondest delusions and it will hold you back from truly studying anything on the topic#yelling at the void
6 notes
·
View notes
Text
sometimes it feels like i'm the only one in this fandom who's right about teddie,,,,,, /silly
#rambearling#persona 4#p4#teddie persona 4#teddie hanamura#i do see a lot of people who understand him well for the record. and i get that he can be a lot#but i also feel like he's severely overhated-#statistically i see more love for him but the teddie haters are vocal and can be very mean about it-#he's just an autistic teenager who has no concept of social cues or boundaries and is just copying what he sees (mainly from yosuke-)#i love yosuke and all but he's not a good role model. especially not for how to act around girls-#i've said it before and i'll say it again i kin teddie and i also relate to him a lot even now. not necessarily in the exacts obv but#autistic teenager who acts immature for his age and doesn't understand social cues and can be kind of annoying? just like me fr fr#teddie hanamura you are the most character ever...........
5 notes
·
View notes
Text
Summer (?) 1992. The final game supplement for the Mayfair Games DC HEROES ROLE-PLAYING GAME was a loose-leaf WHO'S WHO supplement, intended to complement DC's loose-leaf WHO'S WHO update. Only three of the four planned volumes were released before Mayfair lost the DC license. The first volume includes this entry for Legionnaire Brainiac 5, also covered in 2995: THE LEGION OF SUPER-HEROES SOURCEBOOK, which came out months later.
Brainy's game statistics are mostly the same (although this entry gives him 20 more Hero Points), but the special Psychological Instability rules are unique to this version. This wasn't really a problem anymore in the period when this supplement was published, since it really refers to one particular story from 1979 (a guy loses his shit and creates a universe-destroying super-monster ONE TIME and nobody ever lets him live it down …), but it's an interesting mechanic, so I can see why it was included. The reverse side also has a handy checklist of Brainiac 5 appearances, for completists:
The Personality/Role-playing section on the front side makes some dubious assertions. The first paragraph says:
He might be incredibly intelligent, but Brainiac 5 has a great deal of difficulty when it comes to expressing his emotions. When he first joined the Legion, he tried to concentrate on his feelings to fit in with his teammates, none of whom was Coluan. He even fell in love with Laurel Gand and spent a number of peaceful years with her. But inevitably, he became so overwhelmed with leading the team and protecting the universe that he was forced to forgo the luxury of emotion so that he could use his intelligence to its capacity.
While the 2995 sourcebook was written by Legion scripters Tom and Mary Bierbaum, this supplement was not, and I'm not sure what Winninger was talking about in the final sentence. What had happened in the latter third of the 1984–1989 Levitz series was that Brainy had fomented a conspiracy to avenge the death of Superboy by destroying the Time Trapper, one of the Legion's most powerful enemies, a plan that involved sacrificing Brainy's old friend Jaxon Rugarth. The other Legionnaires then put Brainy on trial for violating the Legion code, and although he was exonerated, he was so annoyed by their attitude that he resigned in a fit of pique and went back to Colu. He returned during the Magic Wars and sort of mended fences, although the team subsequently collapsed and he went on to other things during the five-year gap (principally trying to find a cure for the "Validus plague" afflicting Saturn Girl's kids). Some of the specific circumstances were subsequently retconned in ways not reflected in the actual comics, but that was still basically the gist at the time this supplement was published. It was messy, but it was certainly NOT a matter of his forgoing "the luxury of emotion."
The second paragraph says:
As a Brainiac, Dox has an affinity for pure logic. As time goes on, he seems less and less interested in establishing normal relationships with his teammates in the reformed Legion.
This was not at all true. Indeed, one of the charming aspects of the "Five Years Later" period was that Brainy had actually mellowed quite a bit. He was still a little awkward, and there was unresolved baggage between him and Laurel (who in the interim had had a baby with his best friend!), but he was more at peace with himself, and more patient with the people around him, than he'd ever been in past Legion stories. According to the Bierbaums' sourcebook, he was even writing sweet little haiku about his former teammates in his spare time — hardly the action of a cold-blooded logician.
#comics#who's who#mayfair games#dc heroes#brainiac 5#querl dox#legion of super heroes#legion of superheroes#laurel gand#the conspiracy storyline showed brainy being ruthless#but hardly unemotional#the actual game statistics are whatever#the dc heroes game designers tried three times to come up with workable gadgetry rules#none of them successful#and modeling a “super-genius” character like brainy in game terms is tough
25 notes
·
View notes
Note
Will you be leaving Tumblr?
Most likely.
From the looks of it, our data has already been compiled and will be handed over tomorrow (I don't have details, the article is locked behind an account creation pop-up) so there's little that can be done for what's already been posted.
I have very little faith that asking those giants to take out our data is going to lead anywhere. There is no obligation in the contract for them to do so.
And even if I were to opt-out, (if that's even going to work at all, remember how well "opting-out" of Tumblr Live worked?) do I really want to keep giving my engagement to a website that feels free to do that kind of scummy shit behind the backs of their userbase? No warning, no talks, no transparency at all? This sits really wrong with me.
So, unfortunately, I most likely will leave Tumblr.
#look my friends#i'm a little burnt-out on caring about AI at all#it could be a wonderful creative tool if the whole practices surrounding it and how they acquire their data sets didn't fucking reek#i care so little about AI art. nothing will ever compare to something that has been handcrafted.#made from human motion. human imagination. human emotion. AI art will never replace human artists.#no AI will ever be able to replace me to tap into my vision and imagination. i do not feel threatened by AI.#in fact i laugh at the very idea of letting some fancy statistic model threaten me#but as little as i care about that i DO care about people making profits off of MY work without my consent#i will not stand idly by and let them make more money off of my back#as we say in french: ''ils nous mangent la laine sur le dos'' they eat the wool off our backs#and i'm not just gonna willingly let them keep eating the wool off of my back#either they fucking pay me for the data or they lose my presence and engagement
15 notes
·
View notes
Note
ive been on vyvanse like 2-3 years now and like everything u described in ur post kinda never goes away but i dunno. i feel like it’s the only medication that works for me so i kinda just find ways around the side effects. like eating big meals in the morning before they kick in and i lose my appetite. i also am either fidgeting or listening to music like allll the time to compensate for the underestimulated feeling.
also the side effects are a lot to take in at first, so u kinda don’t notice all the ways it is helping you. once you get used to it, it feels more helpful and less burdensome.
..omg ok im glad you mananged to get past the side effects/get used to them and tbh idm the appetite thing bc workarounds are there but...the feeling of dullness/seperation/not being quite "you" or like...just feeling like you're not perceiving the world quite right and everything is more muted etc...that's like hell to me tbh !! i would rather do worse in every other aspect of life than give up the lovely intensity with which i feel things and see things and sense things, so having to choose one or the other is such a terrifying prospect...
i rly appreciate this actually, i'm going to force myself thru it for a maximum of 2-3 weeks and if things still feel so simultaneously jittery and distant i will simply stop taking them. good to know it's still a strong possibility that this side effect just doesn't go away w time ! u seem so resilient for getting used to all this tbh omg
also yea eating big meals seems to be smth ppl are saying ! that and protein shakes/bars so im gonna try that tmrw too :) probably not eating pretty much at all today made things worse, we'll see what tomorrow brings ♡♡♡
#the only thing that matters more to me than everything i love is the faculties with which i love those things#theres no point doing great in neuroscience if i dont love it anymore bc everything no longer has any luster#and love is harder to conjure up. there was just...no music in my head it kind of scared me tbh :')#like i want to love this world and in order to love it i need to see it vividly#i want to love machine learning and in order to do so i need to be able to feel intense emotions abt like...literal statistical models#it feels much harder to do on vyvanse it seems#maybe im overthinking idk i hope yhis side effect goes away and if they dont i will probably switch meds or stop#bittenbyalovebug#omg ur url is adorable#asks
6 notes
·
View notes
Text
i do also think the "In that case, a human being inspired by someone's art style, memorizing some vague notion of it and incorporating it into their own art would also be infringing on copyright" argument is intellectually dishonest. I am against IP laws and copyright on the whole. but that's a false equivalency
#i think the argument could. possibly. be made for a person creating their own model intentionally. but overall. it’s not the same process#an animal is not a statistical model. everything you take in you add something to. lol#there is just a lot of people ‘i am very smart’-ing ‘why do you value something made by a human higher than somethng made by a machine hm?’#first of all thats a nothing accusation. and secondly. a human made that machine.
4 notes
·
View notes
Text
Understanding Different Approaches to Risk Classification
Join us as we explore the exciting world of risk. We will delve into different methods such as percentiles, composite indices, regression modeling, machine learning, multivariate statistical models, decision trees, Bayesian networks, and neural networks.
Demystifying the world of risk assessment: A breakdown of the most popular methods When it comes to risk classification, there are several different approaches that can be used to determine the risk level of an option, entity, or individual. These approaches include percentiles, composite indices, regression modeling, machine learning, multivariate statistical models, decision trees, Bayesian…
View On WordPress
0 notes
Text
What demonic marketer decided to use the term "hallucination" to describe when a large language model produces text that isn't true?
Because that's like saying that your magic 8 ball has a tendency to hallucinate.
#It makes these models sound a lot more sophisticated though#Did Chat-GPT come up with the term to make itself look better?#chatgpt#Large language models#If you ask a magic 8 ball if the sky is blue#It might say yes but that doesn't mean it understands the sky#It's just statistically it answers yes a certain percentage of the time regardless of what the sky looks like
16 notes
·
View notes
Text
I love you open source software I love you 1998-looking install wizard I love you websites with one purpose and no css
#am I sad I have to learn yet another statistical modelling programming language? yes.#however I KNOW the runks of the world are on my side#university-hosted website from 2003 you would never hurt me.#look at all this plaintext. I am healed#looks inside my brain at the various statistical programming languages I barely still remember. one of you boys has got to go.#matlab put your head in the hole#you never belonged here.
8 notes
·
View notes
Text
the average rockstar has a 1 in 3 chance of marrying a model so like, statistically, if Eddie becomes a rockstar then Steve —
#made up that statistic for the joke but seriously like all of them do this it's kinda embarrassing#anyway more model!steve aus please#steddie#steve harrington#eddie munson#stranger things#shsm writes
23 notes
·
View notes
Text
utterly distraught that my phone is dying and i'm gonna have to get a new one ;-; my old s9 is about the maximum width that fits comfortably in my tiny child hands and they just keep getting bigger and bigger every year. i would legitimately shell out for models with nicer cameras etc if they had them in small form factors but the major companies just refuse to sell them to me. hell world
#sometimes i get sad about how i have a special chair model for short people and it's still a little too tall for me#i am five one! this is not even medically significant!#...why is it impossible to find height statistics that aren't split by gender. what if i wanted to know what percentage of just Americans#i think it's like 6%. that's not an absurdly small number!#zero rambles
7 notes
·
View notes
Text
it's almost 5 in the morning why do i hate myself
2 notes
·
View notes