Cluster methods of market forecasting. - page 8
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Honestly, the most basic, there was a good man posting articles here https://www.mql5.com/ru/articles/1489 etc..Thank you. I've seen that article, so I'll probably use something like that.
The rectangle highlights the "movement pattern" (up - down and down through the top), analysing such situations is very similar to analysing betting in poker "bet saising"
The rectangle highlights the "movement pattern" (up - down and down through the top), analysing situations like this is very similar to betting analysis in poker
Specialists are present - good.
So, it was strongly recommended (by my deeply respected SanSanych Fomenko) powerful, and , moreover, free, tool for conducting clustering (and all other statistical processing) R. In addition to it, well-known and popular are (1) MATLAB (which I used), (2) systemSTATISTICA(here, in general, just type clustering aspects), (3) SPSS, that also have their ardent supporters.
But let's return to investigation of the simplest, spontaneous method of clustering - the method ofcandlestick patterns.
To conduct an expert assessment on this problem, it is natural to ask traders who actually trade and use this method:" Does the candlestick pattern method work in most cases (when applied correctly) or not?"
Let's go to the vote
https://www.mql5.com/ru/forum/226526
Sandwich-type constructions with a bearish candle next to a bullish candle can create false positives and mask the main movement.
Total vote: 79
Summary of the candlestick pattern method as a useful clustering method.
So, according to https://www.mql5.com/ru/forum/226526 and the opinions on this thread candlestick patterns in their pure form do not work, requiring the consideration of additional factors (some indicators, historical period of trade, market conditions - trend-flat, etc.).etc.) which is different for different traders, i.e. in its pure form a candlestick pattern is not a cluster which can be used as a reference in trading.
It works / does not work can be said about the trading system, and candlestick analysis is not a trading system.
In fact, I can put a tick for option 1, but I can say that there are a lot of false-breaks in the candlestick analysis, further books on candlestick analysis describe well how to enter the market, almost at the program level, but there are no adequate sets of rules for market exits, profit taking, stop loss policy, etc. So it's not easy to talk specifically about trading by candlestick analysis, but as an analysis in itself it's quite readable and looks interesting.
It's just that if we define a pattern by cluster analysis and find entry point, we should find exit point by the same cluster method, but we all start applying fixed TP/SL, or measure them by Fibo levels or at crossing of muwings or change of parabolic, but it's not the same.
It's just that if we define a pattern by clustering, and that's how we find the entry point, we should find the exit point by the same clustering method, but we all start applying fixed tps/sl, or measure them by Fibo levels, or when the muwings cross or the parabolic changes, but that's all wrong.
It's the right thing to do if such a cluster works.
Generally speaking, I am not an expert in candlestick patterns. That is why I ask others' opinion based on what I am doing. I just brought them up as one of the examples of clustering. I would like to discuss different clustering methods to determine market trends. Further, I would like to discuss the impact of volumes on this.
Here, by the way, is a good lecture on the essence of cluster analysis.
Suggestion: Pastukhov's dissertation discusses such an analysis based on the construction of Kagi and Renko.
https://www.mql5.com/ru/forum/228822/page7#comment_6823186