🧵 ML
raphael at Wed, 11 Sep 2024 04:49:47 UTC No. 16373686
So my FSIQ according my psychologist is 100 and CAIT ICAR60 which correlates with gai. I can show proof if needed. I started studying the intricacies of machine learning and I knew theory since 2019, but couldn't do any useful feature engineering on stochastic time series data. I found the math for linear, kernel, multivariate, and tree boosting is just regression nothing else. It's either linear regression or non linear, either way its regression and it fits to data. It being over-fitting or curve fitting isn't of concern. can anyone explain how kernel regression got its name? its just non linear regression, multivariate regression is just multi dimensional regression and so on. what the fuck? does academia like mental masturbation?
Anonymous at Wed, 11 Sep 2024 04:56:21 UTC No. 16373698
>>16373686
>Certified midwit
Just stop dude. Start a bakery or do something else productive suited to your intelligence level. You are never going to understand these concepts well enough to make a meaningful contribution.
raphael at Wed, 11 Sep 2024 04:57:10 UTC No. 16373701
>>16373698
why dont you explain instead of deflecting nigger
🗑️ raphael at Wed, 11 Sep 2024 04:58:45 UTC No. 16373705
>>16373698
average noble laureate is 140 nigger so im not trying to invent agi with a math equation on a 3d plane that finds local minima
raphael at Wed, 11 Sep 2024 05:03:46 UTC No. 16373709
>>16373698
average noble laureate is 140 FSIQ nigger so no im not trying to contribute i know psychometrics mathmatical ability is g loaded at .77 on stanford binet and old sat is .98 so no go there
quit assuming shit like a white faggot
im not going to make a generative activation function or some algorithm that generates synthetic data agi because its impossible
raphael at Wed, 11 Sep 2024 05:05:10 UTC No. 16373712
>>16373709
gradient and tree boosting seem to be the only way to predict stuff on time series data which is what people are using these days
tinder would use clustering shit to identify useful features and then predict what profiles do best thats how they found out dogs being in your profile is useful for matches (cnns)
Anonymous at Wed, 11 Sep 2024 15:27:16 UTC No. 16374358
>>16373686
hmmm no? kernel regression is not just non-linear regression?
and multivariate regression is probably a better name than multidimensional regression. a variable usually refers to the covariates as opposed to the response, so multivariate means there are many covariates but multidimensional could mean other things like the response is multidimensional or even something entirely different so we just use the more precise name.
The economizz at Wed, 11 Sep 2024 20:38:49 UTC No. 16374828
>>16373686
Ur poo. Kys
Anonymous at Wed, 11 Sep 2024 20:47:52 UTC No. 16374851
>>16373709
Lmao poo in the loo Ranjeet
Raphael at Wed, 11 Sep 2024 20:58:54 UTC No. 16374860
>>16374828
Shut up pajeet
go fuck yourself you street shitting curry faggot
Anonymous at Thu, 12 Sep 2024 17:26:14 UTC No. 16377277
>>16373709
>wypipo live rent free in my head grrrr
Anonymous at Fri, 13 Sep 2024 00:07:51 UTC No. 16377756
>>16373686
Don't you have friends to spam with your schizo babble?
Anonymous at Fri, 13 Sep 2024 14:27:21 UTC No. 16378575
>>16373686
you are literally wasting your time