仕事が完了した
実行時間5 日
依頼者からのフィードバック
Great work! A highly professional developer.
開発者からのフィードバック
Very considerate and give you the best offer.
指定
Upgrade Existing MT5 CatBoost Angle EA → Add Self-Learning Engine + Multi-Symbol + Prop-Firm Risk Layer (with 5-Year High-Quality Tick Backtests)
Project Overview
I already have a fully functional MT5 EA based on a CatBoost Angle Strategy (3-model ensemble).
This project is NOT about rewriting from scratch.
I need an experienced Senior MQL5 Developer to upgrade and extend the EA’s architecture by integrating:
A Self-Training Engine (pattern-score reinforcement learning)
Multi-Symbol independent learning (6 symbols)
A full Prop-Firm Challenge Risk-Control Layer
High-quality (99% modeling) tick-data backtests for 5+ years
Clean, modular OOP architecture
CSV-based model persistence per symbol
All CatBoost logic, angle features, and ensemble prediction must be kept exactly as they are.
Your job is to extend the EA with new modules and integrate them cleanly.
Scope of Work
1. Integrate the Self-Training Engine (Reinforcement-Style, MT5 Internal)
No Python needed.
All learning is inside MQL5.
Includes:
Feature builder
Feature binning → pattern hashing
Pattern table per symbol
Online score update (win/loss smoothing + decay)
MAE/MFE tracking
Model persistence (CSV)
Learning must run in parallel with CatBoost, not replacing it.
Final Decision = CatBoost Prediction + Pattern Score + Risk Layer
2. Multi-Symbol Learning Engine
Must support the following symbols independently:
XAUUSD
EURUSD
GBPUSD
USDJPY
US30
DAX40
Each symbol has its own:
Pattern table
MAE/MFE
Model CSV
Score history
Learning cycle
No cross-contamination allowed.
3. Prop-Firm Challenge Risk Layer
Must include:
Account-Level
Daily loss limit
Max drawdown limit
Equity guard
Account shutdown mode
Execution Level
Spread filter
Slippage filter
News/time blocked windows
Min seconds between trades (anti-HFT)
Max lots per symbol
Max total exposure
Strategy-Level
Swing mode (minimum holding time)
Risk Layer must wrap the trading engine:
CatBoost → Self-Training → RiskEngine.AllowNewTrade → Execution
4. High-Quality Tick-Data Backtests (Mandatory)
Modeling Quality: ≥ 99%
Symbols (≥ 5 years each):
XAUUSD
EURUSD
GBPUSD
USDJPY
US30
DAX40
Timeframes:
M15
H1
Files required per backtest:
HTML report
Equity curve PNG
Monthly returns table
Raw trade history CSV
Tick-data source info
Modeling quality screenshot
5. Architecture Requirements
Must follow a clean module design:
[ Prop-Firm Risk Layer ]
↓ allow/deny
[ CatBoost Predictions ]
↓
[ Self-Training Engine ]
↓ Feature Builder
↓ Pattern Hash
↓ Pattern Table (per symbol)
↓ MAE/MFE Tracking
↓ CSV Persistence
No monolithic OnTick code.
Clean classes required.
Deliverables
Updated EA (.mq5 + .ex5)
Self-Training Engine (integrated with existing CatBoost logic)
Multi-Symbol learning engine
Prop-firm-safe risk layer
Model persistence per symbol
All 12 required backtests (6 symbols × 2 TF × 5 years)
Documentation
Clean, extensible OOP architecture
Project Overview
I already have a fully functional MT5 EA based on a CatBoost Angle Strategy (3-model ensemble).
This project is NOT about rewriting from scratch.
I need an experienced Senior MQL5 Developer to upgrade and extend the EA’s architecture by integrating:
A Self-Training Engine (pattern-score reinforcement learning)
Multi-Symbol independent learning (6 symbols)
A full Prop-Firm Challenge Risk-Control Layer
High-quality (99% modeling) tick-data backtests for 5+ years
Clean, modular OOP architecture
CSV-based model persistence per symbol
All CatBoost logic, angle features, and ensemble prediction must be kept exactly as they are.
Your job is to extend the EA with new modules and integrate them cleanly.
Scope of Work
1. Integrate the Self-Training Engine (Reinforcement-Style, MT5 Internal)
No Python needed.
All learning is inside MQL5.
Includes:
Feature builder
Feature binning → pattern hashing
Pattern table per symbol
Online score update (win/loss smoothing + decay)
MAE/MFE tracking
Model persistence (CSV)
Learning must run in parallel with CatBoost, not replacing it.
Final Decision = CatBoost Prediction + Pattern Score + Risk Layer
2. Multi-Symbol Learning Engine
Must support the following symbols independently:
XAUUSD
EURUSD
GBPUSD
USDJPY
US30
DAX40
Each symbol has its own:
Pattern table
MAE/MFE
Model CSV
Score history
Learning cycle
No cross-contamination allowed.
3. Prop-Firm Challenge Risk Layer
Must include:
Account-Level
Daily loss limit
Max drawdown limit
Equity guard
Account shutdown mode
Execution Level
Spread filter
Slippage filter
News/time blocked windows
Min seconds between trades (anti-HFT)
Max lots per symbol
Max total exposure
Strategy-Level
Swing mode (minimum holding time)
Risk Layer must wrap the trading engine:
CatBoost → Self-Training → RiskEngine.AllowNewTrade → Execution
4. High-Quality Tick-Data Backtests (Mandatory)
Modeling Quality: ≥ 99%
Symbols (≥ 5 years each):
XAUUSD
EURUSD
GBPUSD
USDJPY
US30
DAX40
Timeframes:
M15
H1
Files required per backtest:
HTML report
Equity curve PNG
Monthly returns table
Raw trade history CSV
Tick-data source info
Modeling quality screenshot
5. Architecture Requirements
Must follow a clean module design:
[ Prop-Firm Risk Layer ]
↓ allow/deny
[ CatBoost Predictions ]
↓
[ Self-Training Engine ]
↓ Feature Builder
↓ Pattern Hash
↓ Pattern Table (per symbol)
↓ MAE/MFE Tracking
↓ CSV Persistence
No monolithic OnTick code.
Clean classes required.
Deliverables
Updated EA (.mq5 + .ex5)
Self-Training Engine (integrated with existing CatBoost logic)
Multi-Symbol learning engine
Prop-firm-safe risk layer
Model persistence per symbol
All 12 required backtests (6 symbols × 2 TF × 5 years)
Documentation
Clean, extensible OOP architecture
応答済み
1
評価
プロジェクト
22
9%
仲裁
6
33%
/
50%
期限切れ
1
5%
取り込み中
2
評価
プロジェクト
8
0%
仲裁
3
33%
/
67%
期限切れ
4
50%
暇
3
評価
プロジェクト
1
100%
仲裁
3
0%
/
100%
期限切れ
0
暇
4
評価
プロジェクト
150
41%
仲裁
3
33%
/
33%
期限切れ
1
1%
取り込み中
5
評価
プロジェクト
3
33%
仲裁
2
0%
/
100%
期限切れ
0
暇
6
評価
プロジェクト
47
28%
仲裁
14
21%
/
64%
期限切れ
1
2%
多忙
7
評価
プロジェクト
0
0%
仲裁
0
期限切れ
0
暇
8
評価
プロジェクト
0
0%
仲裁
1
0%
/
100%
期限切れ
0
暇
9
評価
プロジェクト
18
28%
仲裁
0
期限切れ
3
17%
暇
10
評価
プロジェクト
1
100%
仲裁
0
期限切れ
0
暇
11
評価
プロジェクト
2
0%
仲裁
1
0%
/
100%
期限切れ
1
50%
暇
12
評価
プロジェクト
29
21%
仲裁
20
10%
/
50%
期限切れ
8
28%
仕事中
13
評価
プロジェクト
553
50%
仲裁
57
40%
/
37%
期限切れ
227
41%
仕事中
14
評価
プロジェクト
0
0%
仲裁
0
期限切れ
0
暇
15
評価
プロジェクト
2
0%
仲裁
4
25%
/
50%
期限切れ
1
50%
暇
16
評価
プロジェクト
16
13%
仲裁
4
50%
/
25%
期限切れ
4
25%
取り込み中
17
評価
プロジェクト
3
0%
仲裁
1
100%
/
0%
期限切れ
0
暇
18
評価
プロジェクト
13
38%
仲裁
1
0%
/
100%
期限切れ
1
8%
暇
19
評価
プロジェクト
1
0%
仲裁
0
期限切れ
0
暇
20
評価
プロジェクト
0
0%
仲裁
0
期限切れ
0
暇
パブリッシュした人: 3 codes
21
評価
プロジェクト
0
0%
仲裁
0
期限切れ
0
暇
類似した注文
SMC ORDER BLOCK
30 - 60 USD
I want already build FULLY AUTOMATED order block MT5 XAUUSD HTF H4 ENTRY LTF M15 - Show result on live account. m15 ob entry in the direction of h4 ob bias the developper to provide source code in the end
Project Title: Custom XAUUSD Support & Resistance Indicator Platform Required: MT5 preferred. If possible, also provide TradingView Pine Script version later. Main Goal: I want a custom indicator made specifically for XAUUSD (Gold) only. The indicator should automatically detect and draw strong support and resistance zones where price has a high probability of reacting, rejecting, or reversing. It must update
Boom and crash robot
30+ USD
I need a boom-and-crash MT5 robot that is very accurate for opening trades just before the spike happens. It should use M1 timeframe It should have options for changing lot size, number of trades to open, stop loss and take profit in points If the spike happens and it makes some profit, it should automatically close all trades. If the spike happens and the spike is not in profits, the trades can remain open waiting
1. IF price forms: - Higher highs + higher lows → TREND = BUY - Lower highs + lower lows → TREND = SELL ELSE → NO TRADE 2. IF: - Trend = BUY - Price retraces to support zone - Bullish engulfing candle forms - TDI green crosses above red (optional) THEN: - Execute BUY 3. IF: - Trend = SELL - Price retraces to resistance - Bearish engulfing forms - TDI confirms THEN: - Execute SELL 4. Risk per trade = 1% of account Lot
HFT trading robot needed
30+ USD
I need a high frequency trading robot for gold in one or 5 minute timeframe the robot should have spread filter where it should only open trades below a set spread should have news filter to allow trading during fundal news or not the robot should have input in number of minutes to close all open trades and remove pending orders before fundamental news as part of news filter. It should also have the number of minutes
Hello, I am looking for a professional trading system including: 1- Trading Bot (Expert Advisor): - Good profit performance - High security and strong risk management - Works efficiently during high market volatility (news and strong movements) - Works on all pairs (Forex + Gold) 2- Signal Indicator: - Provides clear Buy and Sell signals - Includes Take Profit and Stop Loss - No repaint (signals must not change or
Yakubu Jnr Ai Bots
30+ USD
This is yakubu Jnr trading bots I create the trading robots to help my self and others traders to be successful please you can join my live trading bots or subscribe to my trading robots
Apply with a screen of your work . Symbol Specific Logic . Live Chart Optimization Check the Core logic . [back tests as well] Change points to pips . Create buffer for the zone
Tengo una estrategia basada en divergencia para el oro sobre todo en tf m1 Basado en divergencia con stoch .. confirmando la entrada con ciertos parameteos de entrada Es mejor conversarlo para dar mejor los detalles Cuando entrar, porque o todas las divergencias se debe tomar para entrar en compras o ventas He adjuntado un ejemplo La confrmacion más exacta es el cruce de esos parámetros de stoch edebajo de level de
プロジェクト情報
予算
550 - 700 USD
締め切り
最高 10 日