What is the perspective in your opinion? - page 14

New comment
 
ivandurak, I wonder how the bidding is analysed?
 
Summer:
ivandurak, I wonder how the bidding is analysed?

We calculate the cumulative sum of the last n trades, the closer the result is to a straight line the better.
 
Summer:

now that you are here, it will be interesting to know that the dream of teleportation is not far away from being realized. Here is interesting news (although it is hard to call it news :) ) Teleportation at 10 miles
If you read carefully everything on my links, it's there too. It's very detailed.
 
ivandurak:

Calculate cumulative sum of last n trades, the closer the result is to a straight line the better.


This is the first time I'm hearing the word "cumulative", so it's hard for me to compare this method as an alternative to the usual optimization in the tester =)

I should probably post the code at once =)

 
Zhunko:
If you read everything on my links carefully, it's all there too. It is very detailed.
Interesting texts, although written in the somewhat strange language of the denouncer-discloser. I agree with the author in one thing - with a probability of much more than 50% our world should be seen as a simulation, the result of the work of some algorithms.
 
Ichor:

__Classify Zigzags - I did it, there are results. To be honest I think Topchilo did something similar either filtering the TS signals or classifying Peaks and Troughs, behind which there should be moves greater than X pips. I was happy with the results, but I didn't do it all in MT, so I didn't see it on the chart, but now I plan to continue my research in this direction in MT5.

Maybe the author will see or maybe somebody else who is into classifiers will ask me a question: do you use GA and are there any specifics with it?
 
Ichor:


... a very good example of classification - Fisher's Iris (types and parameters of types) -> define type, by parameters, IMHO suitable for the market (actually reversals. trend, etc. or signal filtering TS), there is a profitable signal with a set of characteristics and loss signal with a set of characteristics, to separate the wheat from the chaff)

Classification of Fisher's Iris, as opposed to the markets - is not a problem for modern NS.

Such a problem is more suitable for testing homemade nets for lice: if it can't classify irises, it's garbage. Or to test neural network experts for their professional skills.

The attached archive contains Fisher's classification of irises with NS embedded in an Excel spreadsheet.

Files:
fisher.zip  5 kb
1...7891011121314
New comment