Machine learning in trading: theory, models, practice and algo-trading - page 2330

 
mytarmailS:

Musk bought bitcoin four days ago, this is great news

You can see on google trends how people are googling for information

And if you compare the "interest in bitcoin" chart with bitcoin price, you can see the correlation with the naked eye and quantitative correlation > 0.6

Blue bitcoin price, krvsny r.trend

I'm sure Twitter analysis looks the same...

It's not a bad sign...
 
denis.eremin:

It's not clear without a concrete example - show it on pairs

By analogy with randoms, the correlation between dollar crosses should be 0.5. For example between the logarithm increments of eurusd and gbpusd

 
mytarmailS:

On google trends you can directly see how people started googling info

And if you compare the graph "interest in bitcoin" and bitcoin price, you can see the connection with the naked eye and quantitative correlation > 0.6

Blue bitcoin price, krvsny r.trend

I'm sure Twitter's analysis looks the same...

It's not a bad sign...

Musk was buying in january the acid stoner told this

 
Идея, взятая из физики, помогает ИИ работать в высших измерениях
Идея, взятая из физики, помогает ИИ работать в высших измерениях
  • habr.com
Компьютеры учатся водить автомобили, обыгрывают чемпионов мира в настольные игры, и даже пишут прозу. По большей части революция ИИ зиждется на возможностях одного типа искусственной нейронной сети, схема работы которой вдохновлена связанными друг с другом слоями нейронов в зрительной коре мозга млекопитающих. Так называемые «свёрточные...
 

Find 10 differences, so to speak.)

it turns out the bitcoin is completely free, according to pure market principles, without market maker

the correlation is 0.7222562



And if we normalize the series according to volatility it turns out that the bitcoin never fell. It is quite interesting to look not at the absolute price, but at the relative


 
mytarmailS:

... it turns out that the bit has never fallen.


Bull's-eye!

 

By the way, there is a strong suspicion that the series normalized by volatility is better predicted


 

No one cares about the question of why the more money in the system the more expensive bitcoin is?

The math in the pyramid is interesting-interesting.

Soon all the peeps in the world will enter the system, deposit all the money, and paradise will come). Bitcoin will be worth trillions of bucks. And everyone will be able to buy their own factory, where robots will work, and cars on autopilot will deliver products, while the peeps will lie on the beach and buy bitcoins again. The price of a dollar invested will increase many times over in bitcoin.

Isn't that cool? But I'm reminded of tulips;)

 
Uladzimir Izerski:

Bitcoin will be worth trillions of dollars. And everyone will be able to buy their own factory with robots, cars on autopilot will deliver products, and people will lie on the beach and buy bitcoins again. The price of a dollar invested will increase many times over in bitcoin.

So far, it looks that way. those who are not on board remain overboard

 
Rorschach:

By analogy with randoms, the correlation between dollar crosses should be 0.5. For example between the logarithm increments of eurusd and gbpusd

With this knowledge only at the DSP.

Reason: