Machine Learning : self optimizing EA based on 3 time frames, 6 indicators

MQL5 专家

指定

HI,


As you can see, I trade by analyzing the market dynamics on 3 different Timeframes (here H1, 15M, 50T), and 6 different indicators : 

Heiken Ashi, RSI, ATR, Stochastic, Parabolic SAR, Pivot Points (5 timeframes: 1M 1W 1D H4 H2, making the total 10 indicators)


My strategy basically : according to the longer term trend (H1 and 15M), I can define a direction for my short erm trade (50 ticks).

- The strategy aims to take advantage of the short term rebounds going in the direction of the long term trend. 

- When it's possible, take advantage of the long term rebound to achieve bigger returns.

- A rebound occurence is the same pattern on every time frame : 

- price is getting close to (or touches) a Pivot Point level.

- RSI is getting close (or beyond) the tension zone (30 or 70)

- Stochastic is getting close(or beyond the tension zone (20 or 80)

- heiken ashi candles show a slowdown in trend, by getting smaller and smaller, or turning green from red, or red from green. 

- All those parameters before can vary according to the volatility in the market, that’s the level of ATR. 


Here is what I need: 

  • for each time frame (here H1, 15M and 50T), apply this reasoning explained above
  • make predictions for the shorter term TF (predictions based on H1 will apply on 15M and 50T, predictions on 15M will apply on 50T only)
  • take trades automatically in 50T time frame using the same strategy explained above. 


Every part of every indicator has to be optimized by the EA, except for the Pivot Points which are fixed. For exemple in the RSI : on each timeframe, the period used ( default :RSI(14) ), and the low/high points (default : 30/70 ) will be optimized. 

The (trailing) stop loss, position size, and target profit, are also to be optimized by the EA according to what it has learned from the backtests ans its past trades : the position would be smaller if the risk/reward ratio is weaker, and the target will be farther if the prediction system thinks the upside is big, for example. 


Feel free to contact me for further details, 


Thank you

反馈

1
开发者 1
等级
(69)
项目
93
34%
仲裁
10
40% / 30%
逾期
5
5%
空闲
相似订单
Hey guys, I’ve got an MT4 HFT EA that’s showing strong results on an IC Markets demo account, and I'm looking to move it over to MT5. Does anyone have a solid MT5 version that performs well on IC Markets? Having the source code (MQL5) included would be awesome. Thanks for any leads
Hello! I want to automatically trade on Sierra Charts (IronBeam Brokerage). NQ ($20 per "handle") or MNQ ($2 per point or handle) - futures Broker: IronBeam mostly for the low capital requirement for futures Chart: Sierra Charts I almost exclusively trade a 5-20 bracket with 2 contracts (first contract closes at a +5 and second bracket closes at a +20.) When the +5 is reached, I move my stop from -15 to BE
Let’s use Chart M3. Stochastic 8,3,3. 85/15 trigger for sell stochastic >85, and the reversal move need to engulf the last bar for 1 point. SL over the engulfed bar. Same trigger for buys… Stochastic <15, engulfed bar by 1 point = entrance. SL under the engulfed bar… more details as soon we start to work on it
Scalping EA/straddle 30 - 200 USD
Title: Custom EA: Dynamic Trailing Limit Grid with Volatility/Velocity Filter and Auto SL/TP Description: I am looking for an experienced MQL5 programmer to develop a high-performance, robust Expert Advisor (EA) for MetaTrader 5. The strategy relies on dynamic trailing limit orders that automatically adjust to market price changes, featuring a safety buffer during high-velocity price spikes. The EA must include the
Freelance Task — EA Recovery Layer Hello. I have a ready-made MQL5 EA. The signals, SL/TP and lot calculation should essentially not be touched. I need a recovery system added. I am not a programmer — below is how the robot should behave. What stays Entry only from an EA signal. Stop-loss is not a new signal. Trailing is already built into my EA — recovery does not implement its own trailing. If trailing is disabled
Watcher position 30 - 100 USD
//+------------------------------------------------------------------+ //| synthetic tick generator based on two assets | //+------------------------------------------------------------------+ class CSyntheticTickGenerator { private : string symbol_a; string symbol_b; string synth_name; double ratio; double last_price_a; double
# SNRZ BOT — SESSION PROGRESS LOG Last updated: 2026-09-16 (local late-evening session) ## SAN BOT (new — cloned from SNRZ) - `MQL5\Experts\AutomatedOrderBot_SAN.mq5` / `.ex5` — v1.0, compiled 0 errors/0 warnings - Signal source: custom indicator "SAN" (InpSANName, buy buf 0 / sell buf 1, target bar 1), same SNRZ-style read, MA fallback if missing - Magic number 234567 (separate from SNRZ 123456 so both bots can run
Multi-Asset EA (XAUUSD / FRA40) with Time Filters, High-Impact News Avoidance, Advanced Risk/Lot Sizing, and H1 Breakout Strategy 1. RISK SETTINGS * Max Daily Loss Limit: (bool) Toggle true/false. If true, input double for max loss in currency (e.g., $500). If hit, close all positions and stop trading until next day. * Lot Sizing Mode: Dropdown menu: 1. Fixed Lots (e.g. 0.10); 2. Percentage Risk (%) based on Account
Should be as it but no alarms, also we need to add a deep search so he can see all the preves candles and a memory function so its not slowing down the indicator. for higher time frames daily 4h 1h around 365 days and for lower time frames we do up to 20k bars but both flexible so i can change it later. the memory is important so the indicator is not always recalculating everything when a new candle shows especially
Creating a bot using the Parabolic indicator with a simple interface.. I want a bot that uses the Parabolic SAR indicator only, with these orders attached to it. Boot name : HEMIN .FX Magic number : 000000 …………………………………………………. Trade behavior : regular OR inverse OR only long "buy" OR only short "sell" Timeframe : ………… Period: ……… ………………………………………………….. Lot size : ……………… Break even in points : …

项目信息

预算
30 - 70 USD
截止日期
 10 天