Работа завершена

Время выполнения 5 дней
Отзыв от заказчика
Great work! A highly professional developer.
Отзыв от исполнителя
Very considerate and give you the best offer.

Техническое задание

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

Откликнулись

1
Разработчик 1
Оценка
(21)
Проекты
27
7%
Арбитраж
9
33% / 33%
Просрочено
1
4%
Работает
2
Разработчик 2
Оценка
(4)
Проекты
8
0%
Арбитраж
3
33% / 67%
Просрочено
4
50%
Свободен
3
Разработчик 3
Оценка
(3)
Проекты
1
100%
Арбитраж
3
0% / 100%
Просрочено
0
Свободен
4
Разработчик 4
Оценка
(169)
Проекты
180
46%
Арбитраж
3
33% / 33%
Просрочено
1
1%
Работает
5
Разработчик 5
Оценка
(4)
Проекты
3
33%
Арбитраж
2
0% / 100%
Просрочено
0
Свободен
6
Разработчик 6
Оценка
(48)
Проекты
58
34%
Арбитраж
15
27% / 60%
Просрочено
1
2%
Работает
7
Разработчик 7
Оценка
Проекты
0
0%
Арбитраж
0
Просрочено
0
Свободен
8
Разработчик 8
Оценка
Проекты
0
0%
Арбитраж
1
0% / 100%
Просрочено
0
Свободен
9
Разработчик 9
Оценка
(31)
Проекты
33
42%
Арбитраж
0
Просрочено
3
9%
Свободен
10
Разработчик 10
Оценка
(1)
Проекты
1
100%
Арбитраж
0
Просрочено
0
Свободен
11
Разработчик 11
Оценка
(1)
Проекты
2
0%
Арбитраж
1
0% / 100%
Просрочено
1
50%
Свободен
Опубликовал: 4 примера
12
Разработчик 12
Оценка
(29)
Проекты
33
27%
Арбитраж
20
10% / 50%
Просрочено
11
33%
Свободен
13
Разработчик 13
Оценка
(271)
Проекты
553
50%
Арбитраж
57
40% / 37%
Просрочено
227
41%
Работает
14
Разработчик 14
Оценка
Проекты
0
0%
Арбитраж
0
Просрочено
0
Свободен
15
Разработчик 15
Оценка
Проекты
2
0%
Арбитраж
4
25% / 50%
Просрочено
1
50%
Свободен
16
Разработчик 16
Оценка
(22)
Проекты
29
34%
Арбитраж
4
50% / 25%
Просрочено
5
17%
Работает
17
Разработчик 17
Оценка
(2)
Проекты
3
0%
Арбитраж
1
100% / 0%
Просрочено
0
Свободен
18
Разработчик 18
Оценка
(13)
Проекты
13
38%
Арбитраж
1
0% / 100%
Просрочено
1
8%
Свободен
19
Разработчик 19
Оценка
(1)
Проекты
1
0%
Арбитраж
0
Просрочено
0
Свободен
20
Разработчик 20
Оценка
Проекты
0
0%
Арбитраж
0
Просрочено
0
Свободен
Опубликовал: 3 примера
21
Разработчик 21
Оценка
Проекты
0
0%
Арбитраж
0
Просрочено
0
Свободен
Похожие заказы
A good trend predicting indicator is the one which can identify the trend change as soon as it happens on the chart. when a new candle is formed it should tell whether its going to go up or down. I have already seen a lot of repainting trend predictors so if your indicator is repainting then please don't bother contacting. I would like to see the demo version and then if satisfied , I would want the source code too
Hey, I am looking for an experienced MQL5 developer to build a price-action EA for MT5 - XAUUSD and EURUSD on the 30M timeframe. Features required: Pattern recognition based on 3-candle structures and wick ratios. Proportional dynamic lot sizing based on account equity tiers. Time filters, Friday hard-close safeguard, and MT5 native calendar news filter (CPI/PPI lockout). FTMO risk controls (max spread, daily
Bonjour, je recherche un développeur MQL5 expérimenté pour créer un Expert Advisor pour MetaTrader 5 basé sur une stratégie de trading intégrant des principes de gestion des risques rigoureux et d'intelligence financière. Le robot doit être capable de gérer plusieurs paires de devises et d'optimiser automatiquement les entrées et sorties en fonction de conditions de marché prédéfinies."
MT4/MT5 HFT EA us30 30 - 3000 USD
Hello everybody, I'm looking for an experienced MQL4/MQL5 developer to optimize a High-Frequency Trading (HFT) Expert Advisor for both MT4 and MT5. The EA performs consistently and profitably on demo accounts, but when it is run on Raw and Standard live accounts under what appear to be the same trading conditions, it begins generating losses. I do not have the original source code (.mq4/.mq5); I only have the
8 cap prop firm passing 30 - 3000 USD
I am looking for an experienced MQL4/MQL5 HFT developer to build or optimize a High-Frequency Trading (HFT) Expert Advisor that can successfully pass proprietary trading firm challenges and perform consistently under live trading conditions with brokers such as 8cap or BlackBull Markets . The developer should have proven experience with HFT execution, ultra-low-latency trading, broker execution, slippage, spreads
Hello, I am looking for an existing, fully backtestable MT5 Expert Advisor or EA portfolio designed specifically for passing an FTMO-style prop-firm evaluation. I am not looking for unrealistic profit claims or a visually perfect backtest. My priority is a statistically defensible probability of passing the challenge while respecting all risk limits. Primary objective The system should be designed for the current
I have a High-Frequency Trading (HFT) Expert Advisor for both MT4 and MT5 designed primarily for US30 (Dow Jones Index) . The EA performs consistently and profitably on demo accounts, but when I run it on an IC Markets Raw or Standard live account, it starts generating losses under what appear to be the same trading conditions. At this time, I cannot provide the source code (.mq4/.mq5). I can only provide the
Version document : 1.0 Plateforme : TradingView Langage : Pine Script v6 Type : Indicateur d'analyse et d'aide à la décision (non-exécutant) 1. Présentation du projet Nom du produit ONYX SR V2 — Intelligent Support & Resistance Scalping System Objectif Créer un indicateur TradingView capable d'identifier automatiquement des opportunités de scalping basées sur : supports et résistances dynamiques ; action du prix ;
I want to build a fundamental news trading bot that trade off economic news data, as we know every economic news data released always have effect on the asset associated with it, so this bot will take a trade instantly based on the news data released either to buy or sell, it will come with good money management and also SL and TP target based on price and pips value
TumiiFX 30 - 20000 USD
1. Use two EMAs: 20 and 50. If EMA 20 is above EMA 50 → uptrend (look for buys) If EMA 20 is below EMA 50 → downtrend (look for sells) 2. Wait for a pullback into the area between the two EMAs. - For buys: price must touch or move between EMA 20 and EMA 50 during the last few candles. - For stils: same idea, but in a downtrend. 3. Entry signal: Buy: a bullish engulfing candle in an uptrend after the pullback

Информация о проекте

Бюджет
550 - 700 USD
Сроки выполнения
до 10 дн.