XAUUSD Movement Prediction with ML

Job ID: 40289113

Budget: ₹12,500 – ₹37,500 INR

**Project Title:**
Machine Learning Model to Analyze Price Behaviour Around Moving Average (Forex – XAUUSD)

**Project Description:**

We are building a quantitative research project focused on understanding and modeling **price behaviour around a Moving Average** using machine learning techniques. The objective is to develop a **statistical/ML model that can estimate the probability and confidence of price movement relative to a moving average**.

This project is specifically focused on **XAUUSD (Gold) in the Forex market**, using **1-minute historical data**.

The key requirement is that the **model must analyze price dynamics strictly around the Moving Average**, without relying on additional technical indicators such as RSI, MACD, Bollinger Bands, etc.

We want to understand and quantify how price behaves when interacting with a moving average — for example:

* When price approaches the moving average
* When price touches or crosses the moving average
* When price moves away from the moving average
* The probability of continuation vs mean reversion after these interactions

The goal is to build a **machine learning system that can estimate directional movement probability and confidence scores** based solely on the relationship between price and the moving average.

**Scope of Work:**

The freelancer will be responsible for:

1. **Data Analysis**

* Analyze historical OHLCV market data (1-minute timeframe).
* Study statistical patterns of price interaction with a moving average.

2. **Feature Engineering**

* Extract features derived from the relationship between price and the moving average, such as:

* Distance between price and MA
* Rate of approach toward MA
* Price momentum relative to MA
* Price deviation and mean reversion characteristics
* Interaction events (touch, cross, rejection)

3. **Machine Learning Model Development**

* Build a machine learning model capable of estimating:

* Probability of price moving away from the MA
* Probability of price reverting toward the MA
* Expected movement magnitude after MA interaction
* Possible models may include:

* Gradient Boosting
* Random Forest
* XGBoost / LightGBM
* Deep Learning (optional if beneficial)

4. **Training & Validation**

* Proper train/test separation
* Backtesting methodology to avoid data leakage
* Performance evaluation with appropriate metrics

5. **Output**

* The model should produce a **confidence score or probability** for potential price movement relative to the moving average.

**Important Constraints:**

* The system must rely **only on the Moving Average and price data**.
* No external indicators should be used.
* Focus is on **price dynamics around the moving average**.



**Dataset:**

* Forex market data
* XAUUSD
* 1-minute timeframe
* OHLC + volume

**Ideal Candidate:**

We are looking for someone with:

* Experience in **machine learning for financial markets**
* Strong **quantitative research background**
* Knowledge of **time series modeling**
* Experience working with **trading data or algorithmic trading**

**Deliverables:**

* Clean and well-documented code
* Feature engineering pipeline
* Trained machine learning model
* Evaluation results
* Explanation of methodology

This is a **serious quantitative research project**, and we are looking for someone who understands both **machine learning and financial market behavior**.

If you have experience building **ML models for trading or financial prediction**, please include examples of similar work when applying.