|
10+ Jahre
Erfahrung
|
7
Produkte
|
452
Demoversionen
|
|
216
Jobs
|
0
Signale
|
0
Abonnenten
|
🎱🏓🎷💚☕️
💎 The simplest Neural Network Coding Guide : https://www.mql5.com/en/blogs/post/752324
💎 The simplest DLL for mql5 Coding Guide : https://www.mql5.com/en/blogs/post/753426
💎 Harmonic patterns scanner for MT5 : https://www.mql5.com/en/market/product/51212
💎 Similar price action candlestick patterns indicator for MT5 : https://www.mql5.com/en/market/product/133318
💎 Fibonacci Sonar indicator : https://www.mql5.com/en/market/product/133627
💎 Chart projection indicator : https://www.mql5.com/en/market/product/5569
💎 Youtube : https://www.youtube.com/@lorentzor1938
💎 The simplest Neural Network Coding Guide : https://www.mql5.com/en/blogs/post/752324
💎 The simplest DLL for mql5 Coding Guide : https://www.mql5.com/en/blogs/post/753426
💎 Harmonic patterns scanner for MT5 : https://www.mql5.com/en/market/product/51212
💎 Similar price action candlestick patterns indicator for MT5 : https://www.mql5.com/en/market/product/133318
💎 Fibonacci Sonar indicator : https://www.mql5.com/en/market/product/133627
💎 Chart projection indicator : https://www.mql5.com/en/market/product/5569
💎 Youtube : https://www.youtube.com/@lorentzor1938
Lorentzos Roussos
Besides lucky and unlucky theres easy luck and hard luck .
Let's say you buy a pair , and it goes down (trade 1) VS buying a pair -> it arrives one _Point before your take profit level and then goes down (trade 2). Both those trades hit stop loss .
Now let's say you are optimizing this strategy . The more pointers toward an outcome you leave inside the "virtual environment" for the genetic algorithm the more the genetic algorithm will consider the outcome.I call this making the terrain more "walkable".
So , if you are formulating the reward yourself , and , the reward varies between -1.0 -> +1.0 , are these 2 trades equally bad ? Are they both -1 ? I mean if one day you make pizza and you burn it but the other day you make pizza you taste one piece and its good but your girlfriend steals it from you , breaks up with you and leaves (with the pizza) does that mean you will discard both pizza recipes and start over ?☕️
Let's say you buy a pair , and it goes down (trade 1) VS buying a pair -> it arrives one _Point before your take profit level and then goes down (trade 2). Both those trades hit stop loss .
Now let's say you are optimizing this strategy . The more pointers toward an outcome you leave inside the "virtual environment" for the genetic algorithm the more the genetic algorithm will consider the outcome.I call this making the terrain more "walkable".
So , if you are formulating the reward yourself , and , the reward varies between -1.0 -> +1.0 , are these 2 trades equally bad ? Are they both -1 ? I mean if one day you make pizza and you burn it but the other day you make pizza you taste one piece and its good but your girlfriend steals it from you , breaks up with you and leaves (with the pizza) does that mean you will discard both pizza recipes and start over ?☕️
· 3
Yashar Seyyedin
2023.11.10
Great! Also: Unfortunately what you get as a chart has no integrity by itself. Almost meaningless to optimize. Optimization tool is only good to test how much a trading logic depends on parameters' calibration and evaluate its reliability when market atmosphere changes.
Lorentzos Roussos
2023.11.10
you mean the chart comes in with a preselected parameter , the timeframe
Lorentzos Roussos
here are some numbers from the front page of the popular products for MT5
⚠️The average publication date is 9th December 2022
⚠️78% of products have at least one backtest screenshot
⚠️64% of products have a live signal
⚠️The average duration of the signals is 30 weeks
⚠️The average price is 972 $
⚠️64% of products will increase the price after some sales , or the price is a discount
⚠️Theres almost 61 reviews per product with 202 comments and 3170 demo downloads
⚠️36% is using keywords for ai , neural networks , intelligence gpt etc
⚠️You can get an idea of what the product is from the logo in 8% of cases , and from the title in 19%
Interesting observations :
There's products that have fresh live signals but were not published recently
There's an ea that uses chat gpt (which came out 2023) but showcases a backtest from 2019
There's an ea sporting a funded challenge certificate which links out to a different name than the sellers BUT that name appears to be his/hers employee . The employee also has an mql5.com account and has not posted a review on any of the products . Honorable mention . i got curious 😂😂😂🍻
⚠️The average publication date is 9th December 2022
⚠️78% of products have at least one backtest screenshot
⚠️64% of products have a live signal
⚠️The average duration of the signals is 30 weeks
⚠️The average price is 972 $
⚠️64% of products will increase the price after some sales , or the price is a discount
⚠️Theres almost 61 reviews per product with 202 comments and 3170 demo downloads
⚠️36% is using keywords for ai , neural networks , intelligence gpt etc
⚠️You can get an idea of what the product is from the logo in 8% of cases , and from the title in 19%
Interesting observations :
There's products that have fresh live signals but were not published recently
There's an ea that uses chat gpt (which came out 2023) but showcases a backtest from 2019
There's an ea sporting a funded challenge certificate which links out to a different name than the sellers BUT that name appears to be his/hers employee . The employee also has an mql5.com account and has not posted a review on any of the products . Honorable mention . i got curious 😂😂😂🍻
· 2
Lorentzos Roussos
2023.11.09
Also many of the products that have no "FOMO" now did in the past and got to the top . So that number is higher . So the sellers kinda hit the "fair price" the market is willing to pay for their product , stopped the "next price" and are not planning the next scam . i mean the next ai trading algorithm sorry
Lorentzos Roussos
Market History of product price available as a graph at the bottom of the overview or accessible by a "Show Price History" button under the buy choices . Every price change by the vendor creates a new data point on the graph with a date and a price...
290
Lorentzos Roussos
calendar crosshair MT5 for coding warm up . 😊☕️☕️
get it here https://www.mql5.com/en/forum/452930#comment_48934644
get it here https://www.mql5.com/en/forum/452930#comment_48934644
Lorentzos Roussos
Very useful !
10 copies left at 30$ , next price wan million dollors
😂😂😊😇☕️🌴☀️
10 copies left at 30$ , next price wan million dollors
😂😂😊😇☕️🌴☀️
· 2
[Gelöscht]
2023.09.23
wow
Lorentzos Roussos
Hi! First things first you need visual studio for this . Install it , start it up...
· 6
748
24
Lorentzos Roussos
Beitrag Cyprus DST shifts from 2000 till 2028 veröffentlicht
If your broker is in Cyprus If your brokers server time is in Cyprus's Time Zone You can use the following structures to turn historical bar times or historical ticks to gmt time #property copyright "Source" #property link " https://www.mql5.com/en/blogs/post/753335 " #property version "1...
144
Lorentzos Roussos
In this blog i'm exloring how the local memory operates with regard to a work group (of work items). We create a simple kernel that will export IDs , global id, local id, group id of a work item...
214
Lorentzos Roussos
Hi. Short version : 1.CLGetInfoInteger(kernel,CL_KERNEL_WORK_GROUP_SIZE) ; will give you the number of kernel instances that can run at the same time in the device (or the compute unit) (tested with 1CU gpu only) It will also be the max # of items in a group possible, 2...
213
Lorentzos Roussos
beautiful activation function curve .
The derivative is the activation value , that will cover any extra calculations that keep the input range from -5 to 5
😍☕️ (ps:don't forget the normalization derivative)
The derivative is the activation value , that will cover any extra calculations that keep the input range from -5 to 5
😍☕️ (ps:don't forget the normalization derivative)
· 1
Lorentzos Roussos
default execution vs open CL + GPU .
OpenCL 5x , and the algorithm is garbage still !
The top of the line GPU out there is 90x faster than the one that did this!
Source code here : https://c.mql5.com/3/407/imageBlurTest__1.mq5
Thread here :https://www.mql5.com/en/forum/446275
Thanks to William Roeder
[Wait for the Gif to load 😇]
OpenCL 5x , and the algorithm is garbage still !
The top of the line GPU out there is 90x faster than the one that did this!
Source code here : https://c.mql5.com/3/407/imageBlurTest__1.mq5
Thread here :https://www.mql5.com/en/forum/446275
Thanks to William Roeder
[Wait for the Gif to load 😇]
Lorentzos Roussos
Hi there . First things first , watch the following video : Good , now , this is my first openCL program so there may be issues in terminology etc but the goal is to have the simplest example possible , not only because its helpful but also because that's all i can do for now 😇...
150
Lorentzos Roussos
Header file with all activation functions and all derivation functions from metaquotes .
Dive in : https://c.mql5.com/3/406/VectorActivationFunction.mqh
From here : https://www.mql5.com/en/forum/445076#comment_46333973
☕️
Dive in : https://c.mql5.com/3/406/VectorActivationFunction.mqh
From here : https://www.mql5.com/en/forum/445076#comment_46333973
☕️
Lorentzos Roussos
Warning : this is my first attempt at an autoencoder i've seen a video or 2 on autoencoders so i might butcher this With that out of the way let's pop a beer open and grab the keyboard...
395
Lorentzos Roussos
Read Part 1 : The simplest XOR gate neural network Okay . What do we have : nodes , layers and the net and the feed forward functions . So what does that mean in the spectrum of "a problem...
323
Lorentzos Roussos
XOR Gate example neural network in mql5 , as simple as possible 🍺 I'll assume you know what the XOR gate "problem" is . A quick refreshing pro schematic : The first 2 columns are the 2 inputs and the third column is the expected result from that operation...
692
Lorentzos Roussos
Node trying to separate the yellow class samples from the red class samples (dot=sample) 😂
That means :
a.i'm terrible in math 😂
b.the network believes "if you slap it around it will work" 😂(on a serious note , the display is the "world" the network creates )
back to the drawing board...
(90 features one node)
That means :
a.i'm terrible in math 😂
b.the network believes "if you slap it around it will work" 😂(on a serious note , the display is the "world" the network creates )
back to the drawing board...
(90 features one node)
Eugen Funk
2023.04.01
Nice animation. Btw, you won't be able to separate them as long as you use X and y only as "features".
Lorentzos Roussos
2023.04.01
You mean i can't separate them on a 2 dimensional spectrum . (features are 90) but its one layer deep
I won't be able to answer you but you can imagine that it can't be terrible. On the other hand, here is how I see this mediocre tool: If I'm looking for a profitable strategy or out of simple curiosity I first test on time frames of 15 or 5 minutes. If the strategy works or my curiosity is satisfied I already consider it extraordinary compared to manual trading. Then if the stakes are worth it and the strategy can work in shorter UTs I will probably switch to MT5 or something else. You will therefore conclude that this has never happened to date. The ''RENKO style'' strategy was not a great success. Maybe there are errors on my part or other indicators to add. To have. I no longer have the exact reason for my disappointment with the coding of my strategy in RENKO equivalent but it didn't work like on paper. So whatever the quality of the data, it was not at fault or only partially. I did not have the availability to dig in and better exploit this robot and insist on the RENKO equivalent as a strategy but I will surely come back to it at the beginning of 2024 when I have more availability because on paper it had potential. I will contact you by message for the code.