So Veeky Forums I tried to fit a gaussian process model to the bitcoin price...

So Veeky Forums I tried to fit a gaussian process model to the bitcoin price. It works quite well or CO2 concentration (which is steadily rising with seasonal dips) so I thought it would work for BTC.

If it's right it looks like we are fucked.

Time to cash out?

Other urls found in this thread:

scikit-learn.org/stable/auto_examples/gaussian_process/plot_gpr_co2.html
twitter.com/NSFWRedditVideo

Bitcoin is more like this, foo'

Tell me more about this function user. And could you use different model sand compare the likelihoods

I don't see the correlation between the price of Bitcoin and the CO2 being released into the atmosphere...

Bitcoin could very well crash, but have you ever even looked at charts such as Amazon? Bubble bursts, but that bubble bursting is still a blip into the price rise thats to come. Amazon smashed its bubble prices by far... Anyone who holded was off very very well.

Though I'd be more inclined to hold BCH as BTC's price is the only thing going for it. BCH has utility, BTC doesn't really...

I didn't correlate them. Just used a similar model.

also lightning network

This is pretty autistic. Just because you can find a function that traces every previous movement of Bitcoin (pretty fucking easy to do with enough iterations and modification of the function), doesn't mean it can generalize well. Please learn how a training set and test set work in machine learning, then try again.

>I don't see the correlation between the price of Bitcoin and the CO2 being released into the atmosphere...

A lot of models derived from natural phenomena can be applied to things outside their domain, in fact fluid mechanics and particle behaviour is actually used extensively in market analysis. If OP's not larping this looks pretty bad.

Used this scikit-learn.org/stable/auto_examples/gaussian_process/plot_gpr_co2.html

/thread

How to short the climate change market?

hodl ice

Shouldn't I hold water then, if the ice supply is going up?

your first mistake is fitting a gaussian model on non stationary data. At least use log returns or something... not the raw price series.

GAUSSIAN PROCESS ON TIME SERIES DATA ONLY TAKES THE RECENT DOWN TREND INTO ACCOUNT YOU FUCKING IDIOT

woah Veeky Forums is a lot smarter than I thought, I'll have to try harder

ML goy here. I don't think fitting with gaussian rbf is a good predictive model for stocks or crypto OP. Everything is just too random. However, I do think it coincidentally predicted a likely scenario in this case...

No u retard, you should hold ice. Ice suply is going down. FFS how can you be this autistic

it will per definition revert back to the mean. Take the first difference and fit it again.

I said I want to short the climate change market, not invest in it.

Fair enough , I am autistic..

Fuck off Bernie bro. I bet you think only rural and suburban retards voted for Trump too.

I think on average a lot more rural and suburban retards voted for Trump than everyone else