From theory to practice. Part 2 - page 146

 
secret:

Does this "snake" have a clear algorithm?

There are two variants. I'll call them offline and online variants according to the terminology adopted in the matstat.

1) Offline. We fix the sample, then find the moments of the first global high and low achievements - use them to identify the main link of the coil that divides the sample residues into two parts (in the general case). For each of the remaining parts we repeat the same recursively, only after separating the next link the rest of the sample remains only on one side of it.

If the sample is in the past, then the right part of the snake is interesting, and if in the future, then the left part.

2) Online. Sampling grows step by step, and with each step the snake is rebuilt. It's noticeably more complicated here, so it's lazy to write in detail. First, the snake consists of first pair of different ticks. Then each new tick either continues the last link or creates a new one, and the monotony of lengths and lengths of links is checked. If it is broken, the penultimate pair of vertices is discarded - this can be repeated several times.

 
Aleksey Nikolayev:

There are two variants. I will call them offline and online variants according to the terminology adopted by the matstat.

1) Offline. We fix the sample, then find the moments of the first achievements of global highs and lows - we select the main link of the serpentine, which divides the rest of the sample into two parts (in the general case). For each of the remaining parts we repeat the same recursively, only after separating the next link the rest of the sample remains only on one side of it.

If the sample is in the past, then the right part of the snake is interesting, and if in the future, then the left part.

2) Online. Sampling grows step by step, and with each step the snake is rebuilt. It's noticeably more complicated here, so it's lazy to write in detail. First, the snake consists of first pair of different ticks. Then each new tick either continues the last link or creates a new one, and the monotony of lengths and lengths of links is checked. If it is broken, the penultimate pair of vertices is thrown - this can be repeated several times.

This is a good sign.

Finally a ZZ Low knee has been found in the branch and it starts to grow upwards. In terms of usefulness.

 
Uladzimir Izerski:

This is a good sign.

The branch has finally found its ZZ Low knee and it will start to grow up in the branch. In terms of usefulness.

I'm not that optimistic. What would not be useful, they will try to predict the future with the help of what they have found . And if they fail (which they never will), they will throw it in the trash. It is a dead end. You can use SMA in a useful way and not look for anything else :))) The whole question is what to see "use" in the market.

 
Wizard2018:

I am not that optimistic. Whatever useful things are found, they will try to use what they find to predict the future. And if they fail (which they never will), it's a dead end. You may use the SMA in a useful way and not look for anything at all :))) The whole question is what to see the "use" in the market.

And how should one - predict the future using the undiscovered?

 
Andrei Trukhanovich:

Blockchain is inseparable from the underlying token economy, what you describe is either meaningless or useless, or both.

Why, CBDC does exactly that, "meaningless and useless" for users, but meaningful and useful for issuers)))) Rattlesnake works like that, there is centralized manual issuance, "blockchain" is also conditional, but nothing blooms and smells :))

Andrei Trukhanovich:

Getting a token against another token is real, done, used heavily

Is it really necessary? The very meaning of credit is almost completely lost. Credit is a quick exchange of illiquidity for liquidity. If in fact, it is absurd to pledge kefir in exchange for usdt, with such a markup; both are equally liquid. And completely illiquid boards "for some reason" no one wants to bail them out, idiots ...

Andrei Trukhanovich:

Lending is impossible in current realities without KYC and at least partial centralization. Getting a token secured against another token is real, done, used to its full potential.

Read about flashloans, it will be very interesting for you as a trader. If you know how to do it, you may get a million dollars every minute (if you know how to do it and the stars align).

now there are full-fledged trustless or conventionally truseless exchanges that allow you to trade like on a normal centralised exchange, a tumbler, a double auction, all the stuff.

defi is a topic i can sink my head into and stay there for half a year. i'm a defi expert by the standards of this forum, but by the standards of the defi community - a slightly advanced jun.

I need to read, are there any refined sources on the subject, in a short and to the point aboutdefi and dappin the style of whitepeyper Satoshi? I don't want to filter through someone else's endless marketing, I don't have the time.

 
J.B:

Why, CBDC do exactly that, "meaningless and useless" for users, but meaningful and useful for issuers)))) Rattlesnake works like that, there is centralized manual issuance, "blockchain" is also conditional, but nothing blooms and smells :))

there's no blockchain, from the word go, there's a leger, but decentralisation is an important part of the system that's critical for their consensus to work properly

Do you have to?

Yes! I've used it many times myself

putting kefir in exchange for usdt, with that kind of markup is absurd, both are equally liquid, a substitution of meaning.

You are confused, there is no markup, there is a threshold. i.e. for $1000 Ether you can take $700 usdt, if the Ether price goes down by 30%, i.e. the price of the pledge is close to $700 there will be a liquidation.

The idea is to get the asset back without waiting for liquidation.

If you want 100% of the value of the collateral then it's logical to just make the exchange?

Yes I need to read, are there any refined sources on the subject, to keep it short and to the point inthe style of Satoshi's whitepaper, aboutdefi and dapp mainstreams? I don't want to filter through someone else's endless marketing, no time.

the ether ecosystem for all directions has probably a hundred or so major product services and protocols, i'm not even familiar with half of them, and they are not all


defi ecosystem only:


 
Wizard2018:

I'm not that optimistic. You can use the SMA to your advantage and not look for anything else at all :))) The whole question is what to see "use" in the market.

You are a real optimist, aren't you?) The SMA is also useful))
 
denis.eremin:

How do you predict the future with what you can't find?

Like this.

https://iarduino.kz/blog/robot-dvizhucshijsya-po-linii-s-ispolzovaniem-pid-upravleniya.hml

Робот движущийся по линии, с использованием PID управления
Робот движущийся по линии, с использованием PID управления
  • iarduino.kz
Робот движущийся по линии, с использованием PID управления
 
CHINGIZ MUSTAFAEV:
You are a real optimist)) SMA and with good reason too))

In the hands of the Master and the pencil is a formidable weapon. Masha separates Yin from Yang on the graph no worse, and often much better, than the Multi-storey Integral. And then everyone will benefit from this division proportionally to his or her understanding :)))))

 
Wizard2018:

For example like this

https://iarduino.kz/blog/robot-dvizhucshijsya-po-linii-s-ispolzovaniem-pid-upravleniya.hml

Once again, radio amateurs' songs and dances)

By the way, it's a good exercise for the brain to think about the difference between a track laid by someone and a price chart)