The PROphet EA consists of 2 independent from each other linear perceptrons.

Each of them divides the input attributes-candestickls shifting into 2 classes.

Perceptron № 1 >

class № 1 : BUY and class № 2: flat or SELL

Perceptron № 2 >

class № 1: SELL and class № 2: flat or BUY

This is the feature of the EA – not putting only BUY or SELL classes into one perceptron !

The optimization is performed for the last 12 weeks, at the weekend in two stages.

S t a g e № 1:

the daBUY=true and daSELL=false variables are set and only the weights are

optimized x1,x2,x3,x4 from 1 to 200, and the preliminary stop-loss slb from 30 to 100, that is the end of the first stage.

S t a g e № 2:

the daBUY=false and daSELL=true variables are set and only the weights are

optimized y1,y2,y3,y4 from 1 to 200, and the movable stop-loss sls from 30 to 100.

Once the optimization is performed both variables daBUY and daSELL are set true.

The obtained values are valid for the next (future) week.

The adaptation of the parameters after every week passed is performed by the method described here.

The results of the forward testing of a “common” week from 21-st of July 2008 to 26-th July 2008:

Strategy Tester Report

PROphet

Symbol		EURUSD (Euro vs US Dollar)
Period		5 Minutes (M5) 2008.07.21 00:00 – 2008.07.25 22:55 (2008.07.21 – 2008.07.26)
Model		Every tick (the most precise method based on all available least timeframes)
Parameters		daBUY=true; x1=9; x2=29; x3=94; x4=125; slb=68; daSELL=true; y1=61; y2=100; y3=117; y4=31; sls=72;

Bars in test	2420	Ticks modelled	46219	Modelling quality	90.00%
Missmatched charts errors	12

Initial deposit	1000.00
Total net profit	145.00	Gross profit	215.00	Gross loss	-70.00
Profit factor	3.07	Expected payoff	18.13
Absolute drawdown	61.00	Maximum drawdown	70.00 (6.11%)	Relative drawdown	6.11% (70.00)

Total trades	8	Short positions (% won)	5 (100.00%)	Long positions (% won)	3 (66.67%)
		Profit trades (% of total)	7 (87.50%)	Loss trades (% of total)	1 (12.50%)
Largest		profit trade	138.00	loss trade	-70.00
Average		profit trade	30.71	loss trade	-70.00
Maximum		consecutive wins (profit in money)	7 (215.00)	consecutive losses (loss in money)	1 (-70.00)
Maximum		consecutive profit (count of wins)	215.00 (7)	consecutive loss (count of losses)	-70.00 (1)
Average		consecutive wins	7	consecutive losses	1

In this approach to the “prophesy” good hedging in the trading is possible and not only inside one day.

Similar MetaTrader Trading Tools:

MetaCOT 2 CFTC ToolBox (Set of… Contents GENERAL PROVISIONS Brief description Installing the libraries and indicators Several indicators in one window Exporting CFTC reports to CSV…
MQL Plus Enhanced Debugging… Features Features three domains of functionality. - Debugging, Tracing and Performance measurments - Simple to use "Print any variable" to…
Gold Dust - EA MetaTrader 5 Idea by Yury Reshetov mq5 code author - barabashkakvn A Bit of Theory Many automated trading systems tend to over-optimize and over-fit results,…
MTC Сombo - EA MetaTrader 5 Below is the author's descriptions along with my minor edits since five digits are already used for currency pairs nowadays.…
PROphet - EA MetaTrader 5 Author of the idea — Boris, author of the MQL5 code — barabashkakvn. Attention: parameters should be optimized on the M5 timeframe for the last…
MACD Sample - EA MetaTrader 5 The MACD Sample EA is included in the standard pack of the MetaTrader 5 client terminal and is an example…
Combo_Right - EA MetaTrader 5 The Expert Advisor is rewritten from MQL4, the author is Reshetov, link to the source How it works There is…
$Creating fractals in MQL5 using the Iterated Function Systems (IFS) - EA MetaTrader 5$
Creating fractals in MQL5 using… Introduction There are many programs, allowing the creation of self-similar sets, defined by Iterated Function System (IFS). See, for example,…